With QE purchases starting a few days ago, some thoughts on how QE works with a nice graph to explain.
Some new thoughts on the “can you make losses printing money” theme, this time with a Hans-Werner Sinn angle.
The situation in Greece continues to tick over. An explanation of the ECB’s decision tonight to cut off various Greek assets from its eligible collateral list.
It’s complicated. But it’s got nothing to do with technocratic ECB rules.
My thoughts on the ECB’s QE announcement. Perhaps a bit long, perhaps a bit disorganised ….
Everyone knows German savers are special, don’t they? Not so much it turns out. Read here.
Last week, the Central Bank of Ireland published an important consultation paper. Despite the boring title (“Macro-prudential policy for residential mortgage lending”) this is a very important initiative that will have profound implications for banking and the Irish property market.
The paper proposes the introduction of restrictions on Loan-to-Income (LTI) and Loan-to-Value (LTV) ratios for Irish residential mortgages. Specifically, it proposes that a maximum of 20 percent of loans should have LTIs over 3.5 and that a maximum of 15 percent of loans should have LTVs in excess of 80 percent. The latter, in particular, would represent a significant departure from the lending standards seen in Ireland over the past few decades.
Given the recent experience in Ireland and elsewhere with property bubbles, I agree that measures of this type need to considered as part of the regulatory policy kit. So, as a general matter, the Central Bank are to be congratulated for being willing to introduce this type of regulation despite its inevitable unpopularity in certain quarters.
In particular, I support the proposed LTI restrictions. These restrictions may be unpopular now with first-time buyers. For example, someone with an income of €50,000 will now be told that the most they can borrow is €175,000. This may appear to reduce their ability to purchase a house if, for example, they had previously planned to borrow €250,000. However, evidence from previous cycles has shown that more generous mortgage credit tends to drive up house prices. This has meant that first-time-buyers end up chasing their own tails, as they all take on additional debt without actually being able to buy better houses. I would be confident that, over time, the LTI limits will restrain house prices and allow first-time buyers to obtain a home without being burdened with huge debts.
That said, I still have some concerns about the way these measures are being introduced and about their timing. In particular, I have reservations about the rapid introduction of an 80 percent LTV standard.
Consultation and Debate
After the global financial crisis, it has become widely accepted that macroeconomic policy-makers need to look beyond using only short-term interest rates to control the economy. Restrictions on LTV and LTI ratios are examples of a set of policy options that have become known as “macro-prudential policy” i.e. policies that are aimed at protecting the stability of the financial system as a whole rather than just focusing on each financial institution separately.
Central Banks, as financial regulators, will generally be the organisations imposing macro-prudential policies. This has caused some debate about the implications of independent central banks setting a much wider range of policies than in the past. Here, for example, are some thoughts from the IMF’s chief economist, Olivier Blanchard.
If you think now of central banks as having a much larger set of responsibilities and a much larger set of tools, then the issue of central bank independence becomes much more difficult. Do you actually want to give the central bank the independence to choose loan-to-value ratios without any supervision from the political process. Isn’t this going to lead to a democratic deficit in a way in which the central bank becomes too powerful? I’m sure there are ways out. Perhaps there could be independence with respect to some dimensions of monetary policy – the traditional ones — and some supervision for the rest or some interaction with a political process.
I don’t claim to have all the answers as to how these new tools should interact with the political process. Still, I do think the Central Bank’s approach appears to be at odds with Blanchard’s ideas. The approach being taken to debate and consultation has been pretty minimalist.
These policies will have a direct effect on the housing decisions of many people and will probably also have important indirect effects. As such, I believe the measures should have been subjected to a longer and more rigorous consultation process, involving a series of public meetings and Oireachtas committees prior to the specific proposals being tabled. As it is, the Bank is announcing a very short consultation process with the plan being to take comments up to December 8 and then introduce very substantial changes to mortgage lending in January.
One of the themes of academic work on macro-prudential policies is that these policies should be used in a cyclical fashion, tightening credit during upturns and allowing it to be loosened during downturns. My reading of the recent international evidence on the use of LTV caps is that they are being used in countries where the authorities are very concerned about the risk of a housing crash, with the implicit policy being to ease these restrictions when this risk is lower.
For example, Canada is a widely-cited example of a country where a set of restrictions on LTVs have been introduced, with maximum LTVs gradually being reduced from 100 percent in 2008 to 80 percent in 2012 (along with various other measures – see this summary from the IMF). However, it is worth emphasising that these measures have been put in place because the Canadian authorities view a large housing crash as a significant possibility. Here is a chart from the latest Bank of Canada Financial System Review showing a steady rise in housing valuations.
The commentary in the report views a housing crash as a potentially severe threat to Canadian banks and the restrictions on lending are being put in place for this reason.
New Zealand is another example of a country that has introduced LTV restrictions, with loans with LTVs above 80 percent now restricted to be no more than 10 percent of the total amount of mortgage lending. Here is a nice paper from the Reserve Bank of New Zealand describing the restrictions. An important message from this paper, however, is that the RBNZ views these restrictions as an explicitly temporary policy.
LVR restrictions are to be used only occasionally, at those points in the financial cycle where there is a real danger of growing systemic risks leading to financial instability. The Reserve Bank does not intend to operate LVR restrictions in a continuous fashion to smooth the cycle, but rather aims to limit the extreme peaks in house price and housing credit cycles.
This raises questions about the current situation in the Irish housing market. Is Ireland now in a situation where, in the absence of these policies, there would be a significant risk of a crash over the next few years?
Residential property prices are certainly growing a very fast pace, up 15 percent in the year ending in August. But, as the picture below shows, prices are still very far off peak levels.
More broadly, valuations look to be in line with the more sustainable levels seen prior to the housing boom. For example, the ratio of prices to rents is back at levels last seen in the late 1990s.
The Central Bank’s new policies are focused on restricting the downside of house price declines by restricting the provision of mortgage credit. However, the current trend of rising prices is not in any way related to easy mortgage lending. Indeed, the Central Bank’s own statistics show that the total stock of mortgage credit continues to decline at about the same pace as it has over the past three years (see the red line below).
Rather than being driven by credit, the evidence points to a shortage of supply as the main factor driving house prices. After the building binge of the Celtic Tiger years, housing completions have slumped. Indeed, given trends in population and household formation, we now appear to be at the point where the previous period of over-building has now been offset by the cumulative under-building of recent years.
Nor does it appear that recent price increases are doing much yet to provoke a supply response. Planning permissions remain at historically low levels. Anecdotal evidence suggests restrictions on the supply of credit to builders as well as a raft of cost-increasing building regulations are at least partly responsible for this lack of supply response.
Against this background, there are reasons to be concerned about the potential impact on the housing sector of the proposed 80 percent LTV requirement in January. As of now, most first-time buyers have been focused on saving about 10 percent of the value of the home. Particularly for those who are also currently paying rent, coming up with this 10 percent is already a challenge. An instant doubling of this requirement would likely force a large number of potential purchasers out of the market, as they would need to wait another few years to double their savings.
This measure may well end up stabilising house prices but it may reflect the deliberate engineering of a nasty equilibrium, one in which Central Bank regulations lead to low demand for new house purchases, matched on the other side by a low level of supply due to credit and other regulatory restrictions. Meanwhile many potential buyers (particularly those unable to sufficiently tap the Bank of Mum and Dad) will feel squeezed out of home ownership and forced to rely on Ireland’s chaotic rental market or its already-inadequate supply of social housing.
At this point, my assessment is that conditions in the Irish housing market do not currently point towards the need for a sudden large change in LTV limits. Whether a gradual introduction of higher LTVs is desirable is a more open question.
LTV versus LTI Restrictions
I noted above that the situations in which countries like Canada and New Zealand have introduced LTV restrictions look quite different from the situation in Ireland today. Indeed, it seems highly unlikely that the authorities in Canada or New Zealand would be imposing 80 percent LTV caps if their property market looked like the current Irish market.
In addition, it is worth noting that LTV restrictions are not a necessary part of the macro-prudential toolkit. The UK has also begun adopting macro-prudential policies aimed at cooling a property market that (in places anyway) shows signs of being genuinely over-heated. However, the Bank of England’s new guidelines have focused solely on LTI restrictions rather than limits on LTVs.
Even in expansions, LTV limits can have unfortunate knock-on effects. Here’s a nice article from the Telegraph on macro-prudential policies. It contains the following observations from Adair Turner.
“The trouble with LTV is it can be a bit circular. You impose an LTV limit and the price can go up, and then somebody can borrow more money via a mortgage on the price that’s gone up,” he says. “One of the problems we have in our economy is the way in which we borrow money against the value of an asset which goes up, which appears to make more borrowing justified. LTI targets the real thing, which is: can people repay the debt out of their income?”
In this sense, a policy framework that focuses too much on LTV restrictions can enhance the fundamental pro-cyclicality that is already in place with existing banking regulations.
Needed: An Integrated Housing Policy
The benefits to financial stability of requiring a large deposit before purchasing a house are clear. Banks are less likely to lose money on home loans in a downturn once the owners have put in more equity. However, the risk associated with falling house prices in Ireland are relatively low at this point and the sudden imposition of an 80 percent LTV norm will have implications for society that go well beyond banks.
It’s not rocket-science to point out that anything that makes it harder for people to purchase something (an inward shift in the demand curve in economist jargon) will result in a lower amount of purchases of that item. The imposition of significantly higher LTVs will likely delay the age at which younger people can purchase homes and will increase the amount of people who will never be in a position to purchase a home.
But housing is special in the sense that we all need a roof over our head, so any policy that reduces people’s ability to purchase homes needs to be matched with a policy that helps them with alternatives such as social housing or long-term renting. Ireland is currently doing a horrible job in this area. The boom provided an opportunity to build up a good stock of high quality social housing but governments decided they had other priorities. Renters also have relatively weak rights, meaning those who want long-term security feel they must own their own home.
These are all issues that the government can act on over the next few years. So I have no problem with Ireland setting a long-term target of an 80 percent LTV norm in the context of a sensible integrated housing policy. But I believe the conditions for the macro-prudential use of LTV restrictions do not apply to the current Irish housing market and that LTI restrictions make better macro-prudential policy tools anyway. As such, I fear the unintended consequences of the LTV policy may outweigh its perceived longer-term benefits.
The concluding lines of the Reserve Bank of New Zealand paper by Lamorna Rogers (cited above) are relevant to the current debate in Ireland.
LVR restrictions provide a way of restraining housing demand while working on the supply response. But in the medium to longer term, imbalances will need to be resolved through appropriate longer run policy measures, including actions to improve the housing supply.
It is, of course, crucially important that Ireland does not return to the irresponsible mortgage lending of the last decade. However, what is required now is more than that: Ireland needs a coherent joined-up housing policy. Ideally, it would be better to debate restrictions on LTVs as part of this broader discussion instead of imposing them for financial stability reasons irrespective of the other parts of the policy framework.
Imagine you’re a smart young rich kid. You invent a money-printing machine. You use it to magic up money to buy $100 in assets. It turns out, though, you didn’t make a great investment and the assets end up being worth only $90.
Do you (a) Be grateful that you’re $90 better off or (b) Call up your Dad to tell him you need $10 to make up your losses?
The debate about potential ECB asset purchases continues to focus incessantly on the issues surrounding potential losses, with the ECB playing the role of the rich kid asking his Dad for an unnecessary bailout.
Take this from my old mate Hans Werner Sinn (via Constantin Gurdgiev).
Aaaand Sinn is baaack… pic.twitter.com/Udldwp8Xev
— Constantin Gurdgiev (@GTCost) October 2, 2014
Apparently tax-payers will have to make up for a shortfall of profits from the ECB if there are losses on ABS purchases. In saying the ECB will have reduced profits, Sinn focuses solely on losing the $10 and forgets about the $90 profit.
One could question whether money creation by the Eurosystem is really just pure profit. Doesn’t this expand the supply of liquidity to the banking system, increasing the amount of money on deposit with ECB and thus increase the interest payments it has to make to banks? Not now – the ECB currently charges banks to keep money on deposit with it so this factor now actually raises the profitability of money creation. (Interest on reserves is also not a necessary feature of a central bank operational framework – the Fed managed fine without it for years.)
I think there are two reasons there is so much confusion about these issues.
The first is the arcane way that central banks do their accounting. All money that is created is counted as a “liability” even if, like bank notes, it doesn’t actually ever impose a cost. Central bank capital measures based on subtracting reported liabilities from reported assets thus grossly underestimate the true financial value generated by central banks. (See here for a more detailed discussion).
Central bank profit and loss measures are calculated in line with this wholly uneconomic methodology. In the analogy above, the central bank would report a $10 loss because its assets had increased by $90 while its “liabilities” had increased by $100. Reporting a loss in these circumstances doesn’t actually make any economic sense but is sure to generate lots of hand-wringing from people who imagine something terrible has happened.
The second problem is the widespread failure to understand that central banks are fundamentally different from commercial banks. Central banks do not need to have assets greater than liabilities and cannot “go bust” due to losses on asset purchases. That said, most of the international policy community seems happy to perpetuate this myth.
For example, consider this Very Serious report from the BIS last year. It recommends that central banks should maintain positive capital, not because they need to, but because the public might be confused about this stuff and so it’s best not to get them too concerned. A quick flavour of the thinking:
we have no doubts about the central banks that are currently shouldering extraordinary financial risks. But our confidence is based on an understanding of the special character of central banks that may not be shared by markets and others.
So if financial markets believed central bankers needed to wear Hawaiian shirts, would it be best to ditch the suits and start drinking pina coladas at press conferences?
There may be another case where an international policy organisation recommends an incorrect policy solely because private investors believe strongly the incorrect policy must be followed, but I can’t think of one offhand.
These points are not intended as a recommendation for central banks to purchases whatever assets just take their fancy. There is an important opportunity cost here. For example, the money could be distributed to the public by providing each person with a fixed amount of money. So it’s important that the purchases are made in a fair and transparent way using market prices.
The other cost usually cited is that money creation may lead to inflation. But in the case of the ECB’s ABS purchases, this is a feature not a bug. The whole point of the programme is to raise inflation back to the ECB’s target level. The fact that the programme also fits with the ECB’s secondary goal of supporting the general economic goals of the EU is welcome but not crucial.
Anyway, expect lots more of this stuff over the next few months as the widespread lack of understanding of central bank balance sheets continues to see irrelevancies held up as crucial economic principles.
When the promissory notes were swapped for long-term bonds held by the Central Bank last year, various people pointed out that the faster the pace of sales of the bonds to the private sector, the smaller the gains from the swap would be. See for example, my post on this and also Seamus Coffey’s similar conclusions. I don’t recall anyone disputing this point at the time.
Now, however, Cantillon is here to tell us that this point is “simplistic”. Apparently, it “ignores the fact that the bank cannot hold the bonds to maturity.” So who is doing this ignoring? Not Seamus. Not me – our calculations on this have always assumed the Central Bank will sell the bonds.
What then about all this stuff about this being a great time to sell the bonds? You wouldn’t know it from Cantillon’s long-winded discussion but this really is a pretty simple issue. If we sell the bonds now, we start paying interest straight away and this adds to the cost for the Exchequer.
Of course, it is possible that yields on Irish sovereign bonds may rise so much in the future that we could end up paying more in interest by delaying the bond sales—we extend the period of time that cost is zero but at the expense of much higher interest costs later and this latter factor ends up dominating. With the ECB committed to low interest rates for the foreseeable future, the Irish economy recovering and the debt-GDP ratio falling, this doesn’t seem like a scenario we need to worry too much about.
Cantillon’s final point – that ultimately what’s going in is that faster sales “diffuses at least some of the anger felt in Frankfurt that Ireland in effect obtained monetary financing for itself via the deal” – is of course spot on. Alas, little is likely to be done to diffuse the anger felt in Ireland about the ECB’s actions during the crisis.
There is lots of good economic news coming out of Ireland these days. GDP is rising at an impressive rate, unemployment is falling, government bond yields are very low and the public finances are improving significantly. In fact, things are so great now that even unmitigatedly bad news is presented as good news.
Thanks to Lorcan Roche Kelly for alerting me to this gem this morning.
.@LorcanRK Wow, a truly impressive amount of wrongness stuffed into one little article. The IT has really set new standards with this one.
— Karl Whelan (@WhelanKarl) September 26, 2014
So why is the Irish Times article so bad? It reports that the Central Bank is speeding up its sales to the private sector of the bonds it received in place of the promissory notes and that it is now going to sell more than the minimum pace of sales signalled last year. The article clearly signals to readers that this is a piece of good news and does not suggest any downside. The reality is that there is no upside whatsoever to the sales. This is a bad news story all the way.
Why is this? The current arrangement features the Central Bank owning bonds issued by the government. The government pays interest on these bonds, these interest payments add to the Central Bank’s profits, and then these profits are eventually recycled back to the government. So as long as the Central Bank holds on the bonds, the net cost of this debt to Exchequer is precisely zero.
What happens when the bonds are sold to the private sector? The annual interest payments now go to private sector investors and don’t get recycled back to the government. So the cost is no longer zero.
The replacement of the despised promissory notes in February 2013 with the new bonds acquired by the Central Bank was widely presented as a big improvement for the Irish state. However, economists emphasised at the time that any benefits depended on the pace of sales. See, for instance, the bottom part of this blog post, which illustrates how a faster pace of bond sales can undo most of the perceived benefits of swapping the promissory notes for longer-term bonds.
But surely there must be some goods news here? What of these capital gains the article refers to? This has occurred because Irish government bond yields have fallen since these bonds were issued to the Central Bank. This means they can be sold for lower yields than the par value they had when the Central Bank purchased them, thus implying a profit on disposal.
The private sector investors who buy these bonds will receive the coupon payments set out in the original bond contracts (they pay Euribor plus 263 basis points) but the capital gain made by the Central Bank means that, on net, the interest cost of the sold bonds to the Irish state will equal the new lower yields. Again, the idea that movements in bond yields would influence the ultimate cost of these bonds was flagged in various discussions of the operation last year, include my own post on it.
So the good news here is the Irish government bond yields have fallen and the net cost of selling these bonds is lower than it would have been a few months ago. But selling them at all still means the cost of these bonds goes from zero to positive: From now on, the bonds are going to have an annual cost to the Exchequer, whereas as long as the Central Bank held them there was no net cost at all. A faster pace of sales thus raises costs for the Exchequer.
Like I said, not a good news story.
Monetary Financed-Related Addendum: One point I omitted when I posted this is the following. Some may read this and say: Why can’t the Central Bank just hand back all of the money it receives from the bond sales to the government, not just the capital gain? The answer is that this would violate the agreement the Central Bank has with the ECB via how to unwind the Anglo situation.
The Central Bank loaned over €40 billion to Anglo\IBRC, most of it in the form of Emergency Liquidity Assistance. This involves the creation of new money. The ECB wanted to see the money issued in this fashion retired from circulation when the loans are repaid. IBRC was liquidated without repaying the loans and selling off the new bonds is the current method the Central Bank has agreed with ECB for how this money is to be retired (or “extinguished” as Patrick Honohan puts it).
So if the Central Bank received a bond with a face value of €1 billion, then a sale of that bond to the private sector should result in €1 billion in money being retired. I’m guessing, however, that if the bond is sold for €1.2 billion because yields have fallen, then the Central Bank gets to keep the additional €0.2 billion and still only retires €1 billion.
Alternatively, all of the €1.2 billion goes towards “extinguishment” but this process could mean we still have some bonds left over (which could be retired from the national debt) once the extinguishing is over. Either way, the €200 million capital gain in this hypothetical example reduces the ultimate net cost of the bond sales but does not change the fact that faster sales are bad news.