Categories
Uncategorized

Ireland’s Credit Guarantee Scheme for COVID-19 SME Lending

The government has announced a loan guarantee scheme for bank lending to Irish SME’s to help them emerge from the economic shutdown associated with the pandemic. The proposed program provides a lending bank with a guarantee giving 80% pari passu (proportional sharing) loan loss protection for eligible loans to Irish SMEs impacted by the pandemic shutdown. There is also a portfolio cap on the guarantee so that each bank can only claim 80% loss coverage on 50% of its covered loan portfolio.  This effectively shrinks the “tail risk” coverage (if the bank’s SME loan portfolio performs disastrously) to 40%. The guarantee is offset by a 50 basis point fee payable to the government. The budgeted €2 billion loan guarantee program equates to 0.58% of 2019 GDP. Policymakers still have a few weeks to best calibrate the program for maximum effectiveness before a prospective Dáil majority coalition passes the enabling legislation. Given the calamitous economic impact of the pandemic shutdown there is no guaranteed best strategy.

It is useful to compare the proposed Irish program with some of the existing programs. Spain has already opened a €100 billion 70-80% pari passu guarantee program; the budgeted magnitude equates to 8.02% of 2019 Spanish GDP. The French €300 billion loan guarantee program equates to 12.4% of French 2019 GDP. It has 90% pari passu loan coverage for lending to firms with 2019 revenues of €1.5 billion or less; 70-80% pari passu for larger firms. Most generous in the EU is the program of Germany, which has announced a loan guarantee program of €500 billion + which corresponds to 14.6% of 2019 GDP. The “plus sign” here denotes that the German government has explicitly committed to increasing the loan guarantee budget to however much above €500 billion is needed. It is not clear whether such an increase is pre-approved by the EU Commission or alternatively whether an increase will require subsequent vetting. The EU amended state aid rules require “the aid is granted on the basis of a scheme with an estimated budget” so it is ambiguous whether the German government can formally make this unlimited commitment within these newly amended rules. The German program pari passu loss coverage ranges from 100% for the smallest firms down to 80% for the largest eligible (the program covers both SMEs and larger firms, but the very largest German corporations are dealt with separately). The German program also offers fast-track approval and dispersal of funds for smaller loan amounts.

The UK is no longer bound by EU strictures regarding state aid rules and monetary financing rules and this flexibility is reflected in its SME lending aid programs. For small loans the UK government offers 100% loss coverage; 80% for larger loans, with no fee for the guarantee. For small firms (less than £41 million 2019 revenues) the government will pay the first six months of loan interest in addition to providing the guarantee, so the implicit “insurance fee” for the guarantee is negative. The small-loan Bounce Back Loan program has had a fast and successful start, whereas the Coronavirus Business Interruption Loan Scheme (larger loan amounts) which requires more vetting and paperwork has had less quick take-up. The UK budgeted amount for its loan guarantee programs is not fixed beforehand. The UK plan originally had a portfolio cap (as in the proposed Irish plan) but they have dropped it.  The very successful joint programs of the US Fed / US Treasury involve 95% outright loan purchases (equivalent to 95% pari passu loss coverage) with no fee payable. One of the two US schemes (the Paycheck Protection Loan Plan, see my earlier blog entry) has a large subsidy component since the loan amount due is partly or entirely forgiven if the loan proceeds are spent on retaining staff that otherwise would have been made redundant. Unlike the US or UK, Ireland has the prospect of a eurozone sovereign debt crisis looming in the background, shrinking the available fiscal space for bold giveaway programs to save jobs. Ireland also must navigate through the state-aid restrictions of the EU Commission.

The economic rationale for these loan guarantees is fundamental and needs to be understood clearly. To encourage a quick macroeconomic recovery, countries need their banking sectors to engage in this lending which, in the absence of a loan guarantee, is not in their financial interest. Sector risks are substantial in this new SME lending since no one knows for sure which currently impacted sectors will remain closed or deeply troubled. For example, there will be likely be considerable SME lending demand from Irish hotels. Should banks be willing to lend to hotels in Ireland, to allow them to reopen? In the absence of a government-funded loan guarantee, the correct answer is no. Commercial banks earn their value by “being boring,” that is lending to low-risk activities, with unsystematic risks which diversify across individual businesses and sectors, resulting in a modest and predictable realised default rate. In exchange the banks earn a relatively small but dependable interest margin over funding costs. The very uncertain prospects for SME lending outcomes in the post-pandemic period, with large systematic sector risks, are too exciting to be a sensible activity for commercial banks in the absence of a government loan guarantee.

The SME loan guarantee programs of EU countries include a 20-100 basis point loan guarantee fee paid to the national government so that each program can be ruled to not violate the EU rules against individual member state aid to industry. It does not make sense if the guarantee fees payable for these programs are market-value based fees which fully compensate for the value of the risk capital. Charging a market-value-based fee for the guarantee defeats the purpose of the program: to give powerful incentives for otherwise-too-risky lending to SMEs in vulnerable sectors.

Patrick Honohan (2020) overviews these loan-guarantee programs internationally and expresses his concern that these loans may impose too much debt on troubled firms and generate prolonged financial distress. Such a concern is particularly pertinent for Ireland, with its extremely slow and cumbersome non-performing loan (NPL) resolution framework. Honohan also worries that many of these government-guaranteed loans may effectively transform into subsidies via non-payment; this is particularly relevant in the case of Ireland given its political-business culture regarding NPLs. Honohan suggests adding an equity-conversion feature to the loan guarantees, but this might be a bit too complicated in the Irish case.

In terms of the Irish proposal, the 80% guarantee coverage is on the low side relative to comparable nations. 90% coverage would be better; imposing a 20% risk exposure on the banks might slow take-up substantially. Getting a fast and high take-up rate requires that the program is administratively easy to access and well-incentivised for both the SMEs and banks in terms of risk-reward acceptability. The €2 billion budgeted amount seems very low. The guarantee fee (which is counterproductive) should be pushed as low as the EU commission will allow. It would be good if the smaller loans at least could have some sweetener attached, linked to payroll or job retention.


Clifford Chance (2020). Coronavirus – Guarantee Scheme in Spain. [online] cliffordchance.com. Available at: https://www.cliffordchance.com/briefings/2020/03/covid-19-_-spanish-guarantee-scheme.html [Accessed 7 May 2020]

Department of Business, Enterprise and Innovation (2020). Credit Guarantee Scheme for COVID-19 FAQs. [online] dbei.gov.ie. Available at: https://dbei.gov.ie/en/What-We-Do/Supports-for-SMEs/COVID-19-supports/Credit-Guarantee-Scheme-COVID-19-FAQ.html [Accessed 7 May 2020]

European Central Bank (2020). ECB announces new pandemic emergency longer-term refinancing operations. [online] ecb.europa.eu. Available at: https://www.ecb.europa.eu/press/pr/date/2020/html/ecb.pr200430_1~477f400e39.en.html [Accessed 7 May 2020]

European Commission (2020). Temporary Framework for State aid measures to support the economy in the current COVID-19 outbreak. [online] ec.europa.eu. Available at: https://ec.europa.eu/competition/state_aid/what_is_new/sa_covid19_temporary-framework.pdf [Accessed 7 May 2020]

European Commission (2020). State aid: Commission approves German measures to support economy in Coronavirus outbreak. [online] ec.europa.eu. Available at: https://ec.europa.eu/commission/presscorner/detail/en/IP_20_653 [Accessed 7 May 2020]

Financial Times (2020). State-backed SME lending picks up pace too late for many. [online] ft.com. Available at: https://www.ft.com/content/9b5c3d0a-4960-4654-b8d5-c4b6697c43c5 [Accessed 7 May 2020]

Financial Times (2020). UK set to launch loans scheme for small businesses. [online] ft.com. Available at: https://www.ft.com/content/52dcf686-6b88-11ea-89df-41bea055720b [Accessed 7 May 2020]

Financial Times (2020). How will the UK’s ‘bounce back’ loans work? [online] ft.com. Available at: https://www.ft.com/content/1ca15db2-93e4-4e34-877d-110309cd9716 [Accessed 7 May 2020]

Financial Times (2020). Loan guarantees: what funding will be available to UK businesses? [online] ft.com. Available at: https://www.ft.com/content/626de1f8-69b4-11ea-800d-da70cff6e4d3 [Accessed 7 May 2020]

Financial Times (2020). More than 100,000 apply for ‘bounce back’ loans. [online] ft.com. Available at: https://www.ft.com/content/ad74d6ff-f9cf-4218-9563-c511681f5acb [Accessed 7 May 2020]

Patrick Honohan (2020). Pandemic loans to firms: Postponing the evil day? [online] Peterson Institute for International Economics. Available at: https://www.piie.com/blogs/realtime-economic-issues-watch/pandemic-loans-firms-postponing-evil-day [Accessed 7 May 2020]

Ireland Strategic Investment Fund (2020). Pandemic stabilization and recovery fund. [online] Available at: https://isif.ie/pandemic-stabilisation-and-recovery-fund [Accessed 9 May 2020]

Bruno Robino (2020). Capped Portfolio Guarantee. European Investment Bank. [online] Available at: https://www.fi-compass.eu/sites/default/files/publications/Bruno_Robino_How_does_a_Guarantee_scheme_work_0.pdf [Accessed 11 May 2020] 

Shearman and Sterling (2020). Updated – Covid-19 France: State Guarantee Scheme for New Money Loans. [online] shearman.com. Available at: https://www.shearman.com/perspectives/2020/04/covid-19-france-state-guarantee-scheme-for-new-money-loans [Accessed 7 May 2020]

The Telegraph (2020). Germany’s 100pc guarantees highlights shortcomings of UK loan scheme. [online] telegraph.co.uk. Available at: https://www.telegraph.co.uk/business/2020/04/08/germanys-100pc-guarantees-throw-spotlight-ailing-uk-loan-scheme/ [Accessed 7 May 2020]

Categories
Uncategorized

Capital Sources for Pandemic Emergency Funding of Irish SMEs: Can Ireland Mimic the US Approach?

The effective closure of the Irish economy due to the pandemic generates very difficult problems in economic analysis; Irish policymakers are struggling to respond quickly. The situation is unprecedented, and it is hazardous to speculate about best policy responses. Nonetheless, with that caveat clearly stated, I want to make some informal remarks about the best ways to get adequate lending to Irish SMEs to help them resume normal business, and the role of the Irish banking sector.

The main point that I want to make is that the best approach to SME support might be through bank-mediated lending in tandem with capital replenishment via loan purchases by the government or central bank. The US Fed has already demonstrated that this works, with a very large loan purchase program already showing positive impact [1]. If loan purchases are not feasible, perhaps some other method of providing contingent capital to the banking sector (to encourage lending) could be used.

It is useful to strip the problem back to some basics: there are three possible sources of funding in this context: government expenditure, private bank capital, and monetary financing through the central bank, and two funding types: cash subsidies or loans. There are of course numerous potential mixtures and combinations of these three capital sources and two funding types.

In the USA, the $349 billion Paycheck Protection Loan Plan (PPLP) is being run by the Small Business Administration in collaboration with the commercial banking system (the $349 billion authorization was quickly exhausted; the amount will likely be topped up this week with an additional $250 billion). Ireland quickly implemented a somewhat parallel scheme, the Covid-19 Pandemic Temporary Wage Subsidy Scheme (TWSS). The PPLP and TWSS have similar objectives, but the TWSS is a direct wage payment/subsidy whereas the American PPLP is packaged as bank-mediated lending with a subsidy attached if the SMEs workers are successfully retained.

The direct-subsidy approach of the TWSS provides a fast start but is limited by its expensiveness per euro of impact. TWSS unlike PPLP also fails to take advantage of the well-developed lending and credit monitoring capabilities of the private banking sector. The Strategic Banking Corporation of Ireland Covid-19 Working Capital Loan Scheme uses the private banking sector but is limited in scope [2].

In the case of SME support based on private bank lending rather than subsidies, it is fair to ask why not rely entirely on private bank capital? Again, it is important to strip back to some fundamental issues. One, the capital at risk from emergency SME lending is potentially large in magnitude and very risky. Two, there is a big public interest in this emergency lending taking place quickly and aggressively to get the economy back up and running normally. The risks are large and the potential (public interest) rewards are also large. The restructuring of the Irish economy post-pandemic could be modest, or it could be massive, and the downturn could be brief or prolonged. Generous SME lending is macroeconomically vital, but risky.

In the USA, the central bank (Fed) quickly implemented a $2.3 trillion debt asset purchase plan backed by its monetary resources. The $2.3 trillion authorized amount equates to 10.7% of 2019 GNP. The Fed is putting a huge amount of risk capital into unusually risky debt assets relative to the classes of assets it has previously had in its portfolio. If this program is successful in helping to restart the US economy, the Fed will get its money back and will have served the national interest. If this risky lending goes sour, which could happen, the risk capital is backstopped by $454 billion (19.7% of the capital amount) that the US Treasury has handed over to the Fed as credit insurance for the program. It is a type of contingent monetary financing which makes good sense under the circumstances.

The Fed purchase program will be split between purchases of private sector debt (78% of the total) and state and municipal debt (22%). As one component of the program, the Fed has stepped in to help facilitate the PPLP; it has launched a $350 billion program to buy up PPLP loans from banks, leaving a residual 5% ownership position in the banks. In tandem with the $350 billion purchase authorization for PPLP-linked loans, the Fed has initiated a $600 billion loan purchasing facility called Main Street Lending Program to purchase non-PPLP bank loans of small and medium-sized US firms. Additionally, the Fed has begun corporate bond purchases of up to $850 billion; note that the US corporate bond market rather than bank lending often serves as a lending vehicle for larger US firms (less true in Europe).

In a rough parallel to the Fed program, the ECB has launched the Pandemic Emergency Purchase Program (PEPP) with authorized funding of €750 billion, which equates to 6.3% of 2019 euro-area GDP. The credit criteria differ from previous ECB asset purchases in that Greek non-investment-grade sovereign debt is included, but there are no major changes to the credit criteria for eligible private debt assets.

One difference between the Fed’s debt asset purchase plan and the ECB’s is that the Fed’s approach is mostly about purchasing private debt assets whereas the ECB’s is mostly about purchasing government debt assets. The ECB’s focus (very understandably) is on preventing a sovereign debt crisis in Italy, Greece and/or Spain; purchasing credit-risky private bank assets is not on the agenda.

Unlike US banks, Irish banks cannot rely on any direct capital support for emergency SME lending from their central bank. Could private bank capital in Ireland prove adequate to fund all pandemic emergency SME lending? Just prior to the pandemic Irish banks had healthy capital ratios and very ample liquidity ratios. Nonetheless, it might be better if these unusual debt assets could be moved off the banking sector balance sheet, as is being done in the US by the Fed’s purchase program. This segregates this unusual lending stream from the other lending activities of the Irish banks and allows them to continue normal lending channels for mortgages, automotive finance, new business finance, and SME expansion. Commingling the normal lending portfolios with this unusual emergency lending is potentially damaging to normal bank lending. Also, if private bank risk capital is used for this SME lending, it does not capture all the public interest rewards from this lending in helping to stabilize the economy. There is a “tragedy of the commons” market failure since the economic gains from a successful lending effort by the banks is shared widely across the economy, but the potential losses associated with the program are paid from private bank capital. This could incentivize banks to under-lend relative to what is needed. Something like the US approach seems appropriate in the circumstances.


[1] See “Federal Reserve takes additional actions to provide up to $2.3 trillion in loans to support the economy” Press Release, Board of Governors of the Federal Reserve System, April 9th, 2020, https://www.federalreserve.gov/newsevents/pressreleases/monetary20200409a.htm; “With $2.3 Trillion Injection, Fedʼs Plan Far Exceeds Its 2008 Rescue” New York Times, April 9, 2020, https://nyti.ms/3caFiH1 (behind paywall) and “Fed Rolls Out $2.3 Trillion to Backstop Main Street, Local Governments,” New York Times, April 9, 2020, By Reuters, https://nyti.ms/3c5qPwk (behind paywall).

[2] See “SBCI Covid -19 Scheme,” Strategic Banking Corporation of Ireland, April 20th, 2020. https://sbci.gov.ie/schemes/covid-19-loan-application

Categories
Uncategorized

A gap in current policies for Irish financial stability

In a recent speech, the Deputy Governor of the Central Bank of Ireland, Sharon Donnery, floated the prospect that the CBI might impose Counter Cyclical Capital Buffers (CCyB) on Irish banks, in order to guard against an unstable credit build-up in the currently strong economic environment. She also used the speech to discuss current conditions in the Irish financial system and review the macroprudential regulation policies of the CBI.

In many ways, Irish macroprudential regulation has been exemplary, but there is a glaring defect. Stanga et alia (2017 and 2018) compare 26 countries regarding mortgage arrears, financial stability and macroprudential policies, and Ireland’s profile is remarkably poor. As Stanga et al. note, controlling mortgage arrears is a key objective of macroprudential policies, and Ireland has very poor performance by this metric.

Ireland’s intractable mortgage arrears problem stems in large part from its defective legal system regarding loan security, with extremely limited lenders’ rights to collateral repossession. This defect in turn limits the reliability of Ireland’s quite restrictive macroprudential policies. As Stanga et al. state in their international overview:

“Better institutions – which improve judicial efficiency and make it easier for banks to enforce their rights – reduce the level of mortgage defaults. We consider several proxies for institutional arrangements and compile an index of institutional quality (IQ). We find a significant and negative relationship between IQ and mortgage arrears, both before and after the onset of the financial crisis – the higher the average quality of institutions, the lower the average mortgage default ratio (Figure 3). Moreover, the effects of macroprudential policies and institutional quality on mortgage defaults are mutually reinforcing. As illustrated in Figure 4, the effect of the MPI [Macro Prudential Index] on defaults becomes stronger in countries with better institutions. This result suggests that the effect of tougher macroprudential policies (that reduce household leverage and ultimately deter defaults) is amplified in an institutional environment conducive to an efficient judicial system with better protection for lenders’ rights and better enforcement capabilities.”

In addition to making banks more cautious, the limited-repossession system in Ireland makes the CBI more stringent in its macroprudential squeeze on credit flows. The prospect of a future spike in mortgage defaults is a key concern for the CBI, along with the high average loss-give-default in such a scenario. Because of this, the CBI is correct to stamp down hard on any signs of substantial credit flow into the domestic housing market.

When it comes to tackling the underlying defect in the Irish system (the too-limited repossession rights of lenders) the CBI has taken the line that this is somebody else’s problem. The CBI harangues the government endlessly on tax and spend policies (which are also not strictly the CBI’s problems) but when it comes to addressing the big defect in the Irish system regarding repossession, the CBI is as quiet as a mouse.

Who is paying for this unusual Irish system of extremely-limited repossession rights? Nondelinquent mortgage borrowers pay for the limited-repossession system since their mortgage interest rate includes the expected cost of default, capturing both a high probability of default and a high loss given default. Households looking for mortgages suffer in two ways: one, the Irish limited-repossession system makes mortgages more difficult to obtain; two, the system has a knock-on effect on housing construction: property development is a high-risk business and with no guarantee of mortgage-ready buyers, developers are extra-cautious.

The net effect of the Irish limited-repossession system on housing prices is indeterminate since there are opposite effects on the demand and supply sides. Cash buyers might benefit or lose on a net basis: they lose from the decrease in house construction (hence higher prices) but benefit from reduced bidding competition against mortgage-based buyers. Existing mortgage holders (other than defaulters) lose, and prospective mortgage holders lose twice over.

At the conclusion of her speech Donnery states:

“While there are uncertainties placing a precise value on the short-term benefits and costs, in the longer-term, increasing the margins of safety in an uncertain world is of benefit to all.”

Consider a young Irish household wishing to buy a family home using mortgage finance. In exchange for a mortgage loan, they might be willing to take a chance that they lose the house in some future scenarios if things turned out badly and they could not pay the loan back. They want a house now and are willing to take a chance on the future. Such a mortgage contract is not legally available to them in Ireland nowadays, since repossession can only be enforced in ridiculously limited circumstances and, due to this legal reality, banks are not allowed to issue mortgage loans unless they are virtually default-risk-free. The young household will have to rent or live with parents, for many years into their future.

The Irish financial system, where there is virtually no chance of receiving a default-risky mortgage and even less chance that such a loan could end with repossession, is not of benefit to all. For many people in many circumstances, risk is good.

Categories
Bailout Banking Crisis EMU European economy

Sharon Donnery (Deputy Governor, Central Bank of Ireland) speech on macroprudential policy

The Department of Economics, Finance & Accounting at Maynooth University welcomes Sharon Donnery, Deputy Governor of the Central Bank of Ireland, who will deliver a talk on “Building resilience in the face of uncertainty – what role for policy?”, followed by a panel discussion, chaired by Bridget McNally (Maynooth University) with panelists Robert Kelly (Central Bank, Head of Macro-Finance Division), Dermot O’Leary (Chief Economist at Goodbody Stockbrokers), and Gregory Connor (Maynooth University), on Thursday 31st May 2018,  at 11am – 12:15 pm, Renehan Hall, Maynooth University. R.S.V.P. EconFinAcc@mu.ie. For further information tel: 01-7083728 / 7083681

Categories
Uncategorized

Four Cheers for Conor Skehan

On Wednesday, Conor Skehan, outgoing head of the Government’s Housing Agency, was grilled by the Oireachtas Housing Committee for the mortal sin of noticing things and speaking honestly about them. Mr. Skehan claimed that some individuals in Ireland were gaming the public housing system, in order to become eligible for public housing ahead of others. Members of the Committee, devout in their observance of the Holy Laws of political correctness,  castigated Mr. Skehan for his public remarks and the evidence he presented to justify them. They noted that it is not possible that a housing-eligible person could game the system – as PC dogma clearly states, lower income individuals are gifted with Immaculate Conception (born without sin) and can do no wrong. So the evidence that Mr. Skehan presented had to be false, and his presentation of it before the committee was proof of his fall from a PC state of grace.

On the plus side, there was at least one honest person in the Oireachtas on Wednesday.

Categories
Banking Crisis Uncategorized

Call for Papers: Fintech and financial risk management: evolution or revolution?

A joint academic-practitioner conference on the theme Fintech and financial risk management: evolution or revolution? will be held in at the Institute of Bankers, Dublin, Ireland on Monday September 10th, 2018. The conference is organized by the Valuation and Risk Cluster (VAR), the Department of Economics, Finance & Accounting at Maynooth University, the Smurfit School of Business at University College Dublin, and the Central Bank of Ireland.

New financial technologies are producing widespread changes to financial markets and financial systems. The effects of the fintech revolution on risk measurement, analysis and control are not yet clear. How does fintech change the risk profile of financial markets? Can existing risk management systems cope with the new environment? What changes are required to existing financial risk management methods and systems? Will innovative applications of fintech improve risk measurement and management

Potential topics include:

• Flash crashes

• Risk measurement and control of black-box trading algorithms

• The impact of high speed trading on dynamic rebalancing and hedging

• Natural Language Processing (NLP)-based artificial intelligence and its trading impact

• High speed trading networks and systemic risk

• Information and noise cascading in networks

• Stability and liquidity of blockchain protocols

• Portfolio risk management with automated advisor systems

• Credit risk in fintech lending systems

• Fintech’s impact on the business models of existing financial institutions

• Applications of machine learning in risk management systems

Please send papers or detailed proposals by May 31st, 2018 at the latest to Na.Li@ucd.ie; all papers must be submitted electronically in adobe pdf format. There will be both main conference sessions and poster sessions. The academic coordinators for the conference are Gregory Connor, John Cotter and Trevor Fitzpatrick, who can be contacted at Gregory.connor@mu.ie,  John.cotter@ucd.ie,  and Trevor.Fitzpatrick@centralbank.ie. The administrative manager for the conference is Na Li who can be contacted at Na.Li@ucd.ie. There are no submission fees or attendance fees for the conference. We are grateful to the Science Foundation of Ireland and the Irish Institute of Bankers for their generous support of this conference. The Valuation and Risk Cluster (VAR) is a collaboration between University College Dublin, Maynooth University, Dublin City University and industry partners, with support from the Science Foundation of Ireland.

Categories
Uncategorized

Honest thoughts on educational inequality in Ireland

In discussing the sources of variation in academic achievement across students, there is a yawning chasm between the contemporary research literature (particularly in the emerging field of geno-economics) and the mainstream media. The mainstream media sticks religiously to the traditional blank slate theory, claiming that variation in student achievement is caused entirely by differing home and school environments. Tuesday’s Education Supplement of the Irish Times is a classic example. The main headline of the supplement is “Privately-educated elite have greater access to education” and the first paragraph reads as follows:

“Young people from disadvantaged backgrounds are denied the same opportunities as their wealthier peers, while parents with money can afford a better education for their children despite Ireland’s so-called free education system, an analysis of the 2017 Irish Times feeder school list shows.”

The article repeatedly relies on the assumption that children of wealthier parents in Ireland perform better in school for only one reason, their parents purchase better educational outcomes through fee-paying schools, tutors, and grind courses. The current scientific literature has an entirely different flavour. A recent paper by Plomin, et al., entitled “The high heritability of educational achievement reflects many genetically influenced traits, not just intelligence,” is typical. Synopsizing their findings, they state:

“Differences among children in educational achievement are highly heritable from the early school years until the end of compulsory education at age 16, when UK students are assessed nationwide with standard achievement tests [General Certificate of Secondary Education (GCSE)]. Genetic research has shown that intelligence makes a major contribution to the heritability of educational achievement. However, we
show that other broad domains of behavior such as personality and psychopathology also account for genetic influence on GCSE scores beyond that predicted by intelligence. Together with intelligence, these domains account for 75% of the heritability of GCSE scores. These results underline the importance of genetics in educational achievement and its correlates.”

To be fair to the Irish Times, an inside piece by Brian Mooney in the Supplement brings a gentle hint of realism into the blank-slate-inspired tirade of the Supplement’s lead article. Mooney hints that there might possibly be other factors explaining why households with two graduate parents grab the university places rightfully going to other households.

“For schools where both parents of many students were graduates, and where they have been supported throughout their education, getting a college place is no great reflection on the success of their school. Alternatively, we are keenly aware that for schools in disadvantaged communities, securing third-level progression for even a small proportion of students is a reflection of highly motivated teachers, and is a fantastic achievement.”

Brian Mooney does not state it explicitly, but scientists have shown definitively that the most powerful “support” that two-graduate-parent households gift to their children is their two tightly packed strands of DNA, which split and recombine, creating a new human infant in the most complex and beautiful physical process in the known universe. This new human infant is not a blank slate; he/she inherits a block-random collection of genomic traits from the maternal and paternal genomes. That genetic process, not fee-paying schools or tutor expenses, is a major source of inequality in educational outcomes in Ireland.

NB: In response to thoughtful comments from colleagues, I changed “the main source” to “a major source” in the last sentence above. That is perhaps more accurate, although it does mess with the rhythm of the final sentence. This edit was made after comments below.

Categories
EMU European economy European politics Uncategorized

Ireland’s Negotiation Game Strategy for Brexit

Ireland needs to play its hand deftly and aggressively during the EU-wide Brexit negotiations. Irish interests in the Brexit process, post-vote, differ from those of other EU states. For EU enthusiasts in states with limited UK trade, a tempting strategy for preventing a NEXT-IT (Netherlands, Austria, Denmark, etc.) is to punish the UK via a spiteful exit deal. That would be a disaster for Ireland due to spillovers. Ireland needs to fight hard to let the UK be allowed a smooth and minimally-disruptive exit, not face a mini trade war. Ireland would be hit very badly in the crossfire.

Categories
Uncategorized

A Brief Reminder: Macroprudential Conference this Friday, January 29th

Details of the macroprudential bank regulation conference this Friday are here. The conference begins at 10 am with an interactive poster session and a chance to meet other attendees.

Categories
Banking Crisis

IMF Post-Program Monitoring Report on Ireland notes the unusual risk profile of the Irish banking sector

The IMF has released its latest post-program monitoring report for Ireland. What is notable in the report is how it highlights the still very-high-risk profile of the Irish banking sector, and the policy quandary regarding encouraging housing construction without endangering another Irish banking sector crash over the medium term. Despite a strong, three-year-long domestic economic expansion, Irish mortgage loans remain an unusually risky asset class.

Categories
Uncategorized

Conference on macroprudential regulation

A conference with the theme Macroprudential regulation: policy dynamics and limitations will be held on Friday January 29th, 2016, at the Institute of Banking, North Wall Quay, Dublin, organized by the Financial Mathematics and Computation Cluster, the Department of Economics, Finance & Accounting at Maynooth University and the UCD School of Business at University College Dublin. The conference schedule appears below.

There are no attendance fees for the conference, but attendees need to register. We are grateful to the Science Foundation of Ireland and the Irish Institute of Banking for their generous support of this conference. The Financial Mathematics Computation Cluster is a collaboration between University College Dublin, Maynooth University, Dublin City University and industry partners, with support from the Science Foundation of Ireland.

10:00 – 10:30  Registration and poster session

10:30 – 10:35  Welcome and introduction

Theme 1: Measuring Macroprudential Policy Needs and Policy Effectiveness

10:35 – 10:55  Marcelo Fernandes (Queen Mary University), “Response to supervisory stress tests”

10:55 – 11:15  John Fell (European Central Bank), “Credit flow restrictions: Implementation and coordination issues”

11:15 – 11:25  Coffee break

11:25– 11:45  Scott Frame (US Federal Reserve), “The failure of supervisory stress testing: Fannie Mae, Freddie Mac, and OFHEO”

11:45 – 12:05  Eugenio Cerruti (International Monetary Fund), “The Use and effectiveness of macroprudential policies: New evidence”

12:05 – 12:25  Panel discussion

12:25 – 13:25  Buffet lunch and poster session

Theme 2: Housing Markets and Macroprudential Policy

13:25 – 13:45  Gabriel Brunneau (Bank of Canada), “Housing market dynamics and macroprudential policy”

13:45 – 14:05  Kieran McQuinn (Economic and Social Research Institute), “Macroprudential policy in a recovering market: Too much too soon”

14:05 – 14:15  Coffee break

14:15 – 14:35  Christoph Basten (Swiss Financial Supervisory Authority), “Countercyclical capital buffers in Switzerland”

14:35 – 14:55  Angus Foulis (Bank of England), “The role of credit in the US housing boom: Insights from tiered housing data”

14:55 – 15:15  Panel discussion

15:15 – 15:20  Closing remarks

Categories
Uncategorized

Philip Lane Appointed Governor of the Central Bank

Congratulations to Philip, and good luck in his important new role.

http://www.irishtimes.com/business/economy/philip-lane-appointed-governor-of-the-central-bank-1.2398674

Categories
Environment European economy World Economy

New York Times: A Migration Juggernaut is Headed for Europe

Eduardo Porter, one of the most highly respected economic analysts in the US media, has an interesting, thoughtful new article on European immigration pressures. He argues that European economies and societies need to prepare for large-scale immigration from Africa, the Middle East, and South Asia. These regions are close to Europe, are notably poor by world standards, and have a forecast population increase of three billion in coming decades, on top of the large increases which have already occurred in the recent past. Porter argues that attempts to stop completely this migration pressure will not succeed, and instead Europe should try to adjust to an inevitable large inflow.

Categories
Uncategorized

Insider or Outsider? Bloomberg article on Honohan succession

There is a brief article in Bloomberg Business today about the search for a new Irish Central Bank governor.

“Ireland is about to deliver evidence on whether, nearly two years after regaining its economic sovereignty, much has really changed. ……   Noonan’s dilemma now is whether to move back to the pre-crisis mode of finding a governor from inside the civil service, or repeat the Honohan recipe and appoint another outsider.”

The Paddy Power betting odds are discussed. As a financial economist, I am forbidden by the Efficient Market Theory from making gambling bets, but perhaps some of the labour/macro economists might want to take a punt.

I had an earlier post on the strategic issues around this appointment. The recent China-related volatility in financial markets is the latest difficult policy problem confronting monetary authorities in Europe and globally. Ireland should appoint someone as governor who can serve both domestically and also contribute to ECB council deliberations.

Since some readers will know some of the people mentioned personally, I blocked the comments feature.

 

Categories
Bailout Banking Crisis Monetary policy Regulation

Macroprudential Regulation Conference: Change of Date

The conference on macroprudential regulation originally scheduled for September 4th has been postponed to Friday, January 29th, 2016. See here for all details on the conference. A full programme will be provided closer to the date.

Categories
Uncategorized

Thoughts on the Design of Grexit

Paul Krugman has a thoughtful op-ed piece in today’s New York Times in which he reluctantly calls for Greek exit from the euro. I share his view on the desirability of Grexit at this juncture, but not his reluctance in expressing that opinion. How could Grexit best be designed, for Greece, for Europe, and for international interests? Below are a few modest thoughts on this.

Categories
Economic growth Uncategorized

African Demographic Trends and European Policy Responses

The Demographic Transition, which started in Europe in the late 18th century, had a huge positive impact on average human welfare. Population levels and growth rates became dependent upon societal preferences rather than upon famine and disease. The demographic transition has now spread around the world to all continents, except Africa. Surprisingly, Africa has not made the switch. Rather than seeing population growth easing and then stopping, in a typical post-demographic transition pattern, African population growth rates have stayed at a very high rate for many decades. Even in recent years, while many demographers expected a slowdown finally to take hold, African population levels have rocketed up. So for example, from the National Geographic:

Categories
Banking Crisis Uncategorized

Hidden Message from the Banking Inquiry: A First-Rate Economist Should Head the Central Bank

Some commentators wrongly claim there is little value in the long and (moderately) expensive banking inquiry. There is much to learn from the inquiry. One important message can be gleaned from the testimony of Central Bank and Financial Regulator executives this past two weeks: the coalition needs to appoint a first-rate economist (like Honohan) as his successor as central bank governor. The coalition should scour the globe and not compromise on analytical firepower.

Brian Lenihan pushed through the appointment of Honohan against the tradition of promoting someone from the senior ranks of the civil service. If the tradition had been followed, the Irish economy might still be wallowing in financial instability. A central bank governor without first-rate economic expertise could have made a total hash of the financial restructuring and recovery programme of the last five years. For example, a former senior civil servant would not have made the phone call to RTE Morning Ireland in November, 2010, getting the Troika programme quickly started. Other painful actions taken in recent years, such as the PCAR and PLAR exercises, and the time-consuming and expensive improvements to the financial sector database, might have never started or been botched. The job requires a highly-competent, well-trained and experienced economist.

Categories
Uncategorized

Irish Financial Regulation: Guidelines Rather Than Rules

In his testimony this morning, Patrick Neary explained that the mandated bank lending sectoral concentration limits, which were seriously breached by the Irish banks during the credit bubble, could not actually be enforced since they were guidelines rather than rules. This distinction between guidelines and rules has eerie similarities to a classic scene in The Pirates of the Caribbean, where pirate captain Hector Barbossa (Geoffrey Rush) explains to Elizabeth Swann (Keira Knightley) the flexibility of The Pirates Code.

Categories
Bailout Banking Crisis EMU Monetary policy

New Working Paper on the Restructuring and Recovery of the Irish Financial Sector

Tom Flavin, Brian O’Kelly and myself have a new working paper on the restructuring and recovery of the Irish financial sector, covering the period late 2008-2014. Helpful comments (cautiously) welcomed.

 

Categories
Bailout Banking Crisis EMU Monetary policy Political economy Regulation

Upcoming Conference on Macroprudential Regulation

Call for Papers: Macroprudential regulation: policy dynamics and limitations

A joint academic-practitioner conference with the theme Macroprudential regulation: policy dynamics and limitations will be held in Dublin, Ireland on Friday September 4th, 2015, organized by the Financial Mathematics and Computation Cluster (FMC2), the Department of Economics, Finance & Accounting at Maynooth University and the UCD School of Business at University College Dublin.

Macroprudential regulation is fairly new, and there are many unanswered questions. Can macroprudential constraints on credit be reliably attuned with the business cycle and/or credit cycle? Are fixed constraints on credit safer and more reliable than attempts at dynamic anti-cyclical ones? Should regulators take account of market or regulatory imperfections, such as in the construction sector, in setting constraints on credit growth? Is macroprudential control by an independent central bank consistent with the democratic accountability of government economic and social policies? Potential topics include:
* Business cycles, financial cycles, and the feasibility of dynamic macroprudential control
* The desirability and effectiveness of LTI and LTV limits on mortgage lending
* Democratic accountability and central bank independence
* Modelling house price movements and household debt and their interactions
* Controlling credit growth and credit flows in the Eurozone
* International case studies of macroprudential regulation.
* Assessment of macroprudential credit-restricting policies

Please send papers or detailed proposals by June 15th, 2015 at the latest to Irene.moore@ucd.ie; all papers must be submitted electronically in adobe pdf format. There will be both main conference sessions and poster sessions. We will consider proposed contributions to the poster session until 31st July. The academic coordinators for the conference are Gregory Connor and John Cotter, who can be contacted at Gregory.connor@nuim.ie or John.cotter@ucd.ie.

There are no submission fees or attendance fees for the conference. We are grateful to the Science Foundation of Ireland and the Irish Institute of Bankers for their generous support of this conference. The Financial Mathematics Computation Cluster (FMC2) is a collaboration between University College Dublin, Maynooth University, Dublin City University and industry partners, with support from the Science Foundation of Ireland.

Categories
Uncategorized

For Risk Measurement Nerds Only: The Swiss Franc Shock was a 200-sigma event

On January 15th, the one-day return to holding Swiss Francs from a Euro perspective was 16.9%. This is a high one-day return for any currency pair, but appears cataclysmic given the extremely low return volatility of the Swiss Franc from a Euro perspective in recent months. This one-day jump was a “239-sigma event” meaning that the magnitude of this return was 239 times the recent return volatility (using a 90-day historical estimate of volatility). In fact, in the period just before the sudden jump, the sample volatility of this exchange rate was even lower. Using a shorter 20-day volatility estimate, the sudden jump was a 400-sigma event.

It is interesting how closely the time-series behaviour of this exchange rate matches the predictions of Krugman’s 1991 model of a government-implemented exchange rate limit, in which traders credibly believe that the authorities will prevent the exchange rate from piercing the exchange rate limit. As the fundamentals for the exchange rate made the Swiss Franc greatly undervalued, the traded exchange rate settled down just near the government-imposed limit, with very low volatility. And then suddenly the credible promise became a non-promise.

Chalk one up for Krugman, in terms of the elegant fit between his theoretical model and this recent market experience. Several forex trading firms went bust, but they should have had better risk management systems.

Categories
Banking Crisis EMU Monetary policy Regulation

Macroprudential regulation: policy dynamics and constraints

The Irish Central Bank is planning to impose macroprudential risk regulation on the domestic banking sector (see here). The general approach of the Irish Central Bank has been widely welcomed by economists, although the specifics of the proposals are controversial.

John Cotter (UCD) and I are planning a conference in September 2015 on macroprudential regulation, the fifth in our series of FMCC conferences on financial risk and regulation. Macroprudential regulation is fairly new, and there are many unanswered questions. Can macroprudential constraints on credit be reliably attuned with the business cycle and/or credit cycle? Are a-cyclical constraints on credit safer and more reliable than attempts at anti-cyclical ones? Should regulators take account of market imperfections, such as the poor performance of the Irish property development industry and the high costs of new housing construction in Ireland, in setting constraints on credit growth?

Macroprudential regulation has particular importance in Ireland, a small open economy buffeted by credit flows from bigger neighbours. The failure to impose macroprudential regulatory control on the Irish banking sector was a central cause of the Irish financial crisis of 2008-2011. During 2000-2007, within a flawed eurozone currency system, a politically-neutered Irish Central Bank ignored a runaway inflow of foreign credit into the Irish banking system. This massive credit inflow undermined the stability of the Irish financial system and led to the disastrous failure of the Irish domestic banking sector.

There is a varied range of views among economists on macroprudential regulation. This is clear in the responses to the Irish Central Bank’s policy discussion document. Three thoughtful responses come from David Duffy and Kieran McQuinn (both at ESRI) here, Ronan Lyons (TCD) here, and Karl Whelan (UCD) here. (For full disclosure, my own response to the Irish Central Bank discussion document is here.) Lyons recommends fixed, a-cyclical credit controls whereas Duffy and McQuinn argue for dynamic, anti-cyclical controls. Duffy and McQuinn stress the need for more new housing in light of fast Irish demographic growth, and the positive role of high housing prices (aided by bank credit growth) in eliciting an adequate supply response. Lyons argues that excessive bank credit growth should not be used as a hidden subsidy for a cost-inefficient building industry.

Lyons makes a case for no loan-to-income (LTI) constraint, instead relying only upon a loan-to-value (LTV) constraint for macroprudential credit control. This contrasts sharply with the view of Karl Whelan who argues for LTI-only macroprudential controls in the current Irish case. Duffy and McQuinn advocate for both controls. I share the view of Duffy and McQuinn. Lyons does not consider the importance of dual-trigger mortgage default in Ireland (that is, mortgage default which is triggered jointly by income stress and negative equity). The amount of Irish mortgage arrears is likely to remain large and volatile, and this is a key potential source of market instability. Both initial LTI and initial LTV ratios are linked to subsequent mortgage default probabilities, so both should be controlled.

There are certainly many points for discussion, which should make for an interesting conference! A formal Call for Papers will follow shortly – if there are particular themes or panels that we should include, feel free to mention them in the comments thread below.

Categories
Banking Crisis Prices Regulation Uncategorized

Composition Effects and Loan-to-Value Limits

The Irish Central Bank is scheduled to introduce new macro-prudential risk controls on Irish mortgage lending, with the new regulations taking effect on January 1st or soon thereafter. One of the regulations will limit most new mortgages to an initial loan-to-value ratio of 80% or less. There has been considerable discussion of the effect of loan-to-value limits on potential property purchasers, but the analysis has been very poorly framed.

The budgeting scenario has been described as follows:

“Consider a couple who wish to purchase a €300,000 property. With a LTV limit of 80% this will require that they save €60,000 for the down payment whereas if they were allowed to borrow 85% they would only need savings of €45,000.”

This oft-repeated budgeting scenario misrepresents the nature of market-wide LTV limits imposed by the Central Bank. This budgeting scenario gives the impression that the policy decision is about imposing/not imposing the LTV constraint on only one particular buyer rather than market-wide. It misses the large compositional effects since leveraged property buyers compete with one another for properties. The degree of leverage allowed in the banking system feeds into property prices, and this affects the opportunity set of purchasers.

Categories
Banking Crisis

NYTimes Op-Ed on Mortgage Limits in the USA

There is an interesting New York Times Op-Ed article relevant to the proposed Irish Central Bank LTV and LTI caps on residential mortgages. US financial regulators attempted to impose very similar caps, but the caps have now been diluted/dropped in response to political pressure.

The article is behind the NYTimes paywall, but a number of articles can be read per month without paying a subscription. A key quote:

“low underwriting standards — especially low down payments — drive housing prices up, making them less affordable for low- and moderate-income buyers, while also inducing would-be homeowners to take more risk.”

Categories
Banking Crisis Uncategorized

Question on measuring foreign risk capital inflows during the Irish financial sector recovery

One of the key drivers behind the better-than-expected recovery of the Irish financial sector has been the strong inflow of foreign risk capital, particularly from U.S. “vulture funds” as they are inaptly named. This healthy demand for Irish banking assets has allowed the PCAR and PLAR plans for the domestic banks, and the unwinding of NAMA, to progress successfully. Similarly healthy demand for the Irish assets of foreign banks, such as Irish loan portfolios sold by Ulster Bank, has also contributed indirectly to the Irish financial sector’s partial recovery.
There is a risk capital inflow when a foreign institution buys a troubled loan portfolio or property portfolio from an Irish bank, or from an Irish subsidiary of a foreign bank, or from Nama. These risk capital inflows are not intermediated through the Irish banks and do not appear on their balance sheets. Prof. Brian O’Kelly (DCU) and I were able to trace the 2000-2009 destabilizing inflow and sudden outflow of foreign credit into the Irish banking sector using the aggregate Irish banking sector balance sheet Table A4.1 published by the Irish Central Bank. Question: how can one measure this new source of risk capital inflows? It seems healthy and stabilizing rather than (like in 2000-2009) unhealthy and destabilizing, but it still deserves to be measured accurately. Is it necessary to list all the individual deals and add them up? Has some hardworking analyst done that already? Is it possible to create a quarterly or annual time series? Answers on a postcard (or better on a spreadsheet) are welcome!

Categories
Bailout Banking Crisis EMU Regulation

The Irish Case for LTV and LTI Caps on New Mortgage Lending

The Irish Central Bank discussion paper on macro-prudential policy tools published yesterday seems to be a trial balloon for possible caps on Loan-to-Income (LTI) and Loan-to-Value (LTV) ratios for new residential property mortgages in Ireland. The general theory behind imposing these limits is laid out clearly in that document; there is no reason to repeat it here. I want to discuss some notable features of the Irish environment which strengthen the case for these caps (but do not make the decision easy).

Categories
Uncategorized

Conference on Financial Crises: Transmission and Consequences

Maynooth University Department of Economics, Finance and Accounting in association with FMC2 (Financial Mathematics and Computation Research Cluster) are hosting a one-day conference on Financial Crises: Transmission and Consequences on Wednesday, September 24 in Renehan Hall, Maynooth University, Maynooth, Co.Kildare.

The event brings together leading international and domestic experts on financial crises, contagion and banking. The full programme of speakers and presentations is shown below. We invite you to join us in Maynooth. Registration is free, but please confirm your attendance by emailing: thomas.flavin@nuim.ie. The conference programme is shown below the fold.

Categories
Uncategorized

Irish Times: New Finance chief told junior official to disregard property crash fears

Interesting short article in today’s Irish Times.

 

 

Categories
Banking Crisis Political economy

The Economic Rationale for the Insolvency Service of Ireland

The economic rationale for the new Insolvency Service of Ireland is well-founded in economic theory. It hinges on the concept of Pareto improving bargains. The idea is that a debtor, with the guidance of a personal insolvency practitioner, can construct a Pareto improving bargain to everyone’s benefit: the debtor, the lender, and society as a whole.

Consider a debtor with unsustainable debt who, to avoid the personal and social costs of bankruptcy, goes to a personal insolvency practitioner (PIP). The PIP objectively examines the debtor’s situation and suggests a payment scheme which offers only part-repayment of loan value. Let the offered proportion of loan value be denoted by OFFER where OFFER < 1.  If OFFER = 1 then the debtor is not insolvent since he/she can afford full-value payment and the PIP has no role.  The PIP describes the offered repayment plan to the lender (or lenders).

The lender knows that the alternative to a personal insolvency plan is bankruptcy for the borrower, and that bankruptcy entails large financial costs, most of which will be borne by the lender. The uncertain proportion of loan value received by the lender after accounting for bankruptcy costs will be denoted by RECOVER.  The debtor will accept the PIP offer if it provides higher expected value of total payments:

OFFER > E[RECOVER],          (A)

where E[ ] denotes the expected value.

The economic rationale for this process is that it can make all three interested parties (debtor, lender, and society) better off. The debtor avoids the personal/social costs of bankruptcy; the lender gets a loan recovery amount which is higher than the expected bankruptcy-cost-adjusted amount received otherwise.  Society avoids administrative bankruptcy costs and gets the benefits of a debtor freed more quickly from debt distress. Of course the PIP has lots of other duties (counselling the debtor, dealing with multiple lenders, administrative duties) but dealing with equation (A) is very fundamental.

The banks understand equation (A); the politicians understood equation (A) when they set up the enabling legislation. Does anyone in the Insolvency Service of Ireland understand equation (A)?  It is fundamental to the Service performing its important task competently.

The Primetime news show recently highlighted a young couple whose PIP offer was rejected.  I do not want to focus particularly on the individual case, keeping in mind the adage “hard cases make bad law.” According to the discussion in the show, the couple owed a mortgage-related debt of €276,000 and their PIP constructed an alternative loan repayment of €2,000. That is, relying on the numbers as discussed in the show, they made an offer of:

OFFER = 2,000/276,000 = 0.0072.

It is important for clarity to note that this does not denote a concessionary interest rate of 72 basis points; rather, 72 basis points is the total proportion of repayment including all principal repayment. Unsurprisingly, the PIP offer was rejected by the lender.

One could argue that the bank could just forgive the couple the loan debt as a gift (skip the 0.0072 partial payment which is too miniscule to consitute a meaningful debt settlement arrangement).  That is, the insolvency system can be brought in as a useful component of parish pump politics, in the good sense, of parish pump politics as using the political system to create unfunded sources of benefits for local causes.  There is certainly a case for doing this, but it was not actually the intention of the legislation. Doing so would greatly increase the effective political power of the ISI as controller of this new source of unfunded social benefits.

A technical feature of equation (A) is a convexifying effect for OFFER proportions close to zero. OFFER is known with certainty whereas RECOVER is a random proportion. Since RECOVER has a lower bound at zero, Jensen’s inequality means that the expected value of RECOVER is much higher than its maximum likelihood value in the region near zero. Is seems extremely difficult to create a scenario where E[RECOVER] could fall as close to zero as 0.0072.

The head of the Insolvency Service of Ireland was on the Primetime show, but he did not seem to be familiar with equation (A), or did not consider it relevant. He did seem to understand that if the ISI had the power to force deals without worrying about (A), then parish pump political considerations would give the agency much greater power. Yet equation (A) was extremely relevant and the absence of any appropriate analysis associated with it detracted considerably from the clarity of the discussion. The staff at the Irish Insolvency Service could benefit from the 30-minute lesson in the economic rationale for their agency’s existence.

[I added a few edits to correct typos, respond to comments (thanks to Sarah Carey and to other commenters who induced me to think more carefully). There may be some time-inconsistencies between the earlier comments below and the later edits.]