PrefBlog

The Bank of Canada has released:

a comprehensive assessment of the potential impact on the Canadian economy of new global capital and liquidity standards, which are to be finalized later this year by the G-20.

The study is titled Strengthening International Capital and Liquidity Standards: A Macroeconomic Impact Assessment for Canada:

While this country boasts a strong financial sector, it too was buffeted by financial shocks from abroad. Our economy could not escape the spillover effects of the ensuing global economic downturn. Canadian economic output fell by more than 3 per cent during the crisis, and more than 400,000 jobs disappeared.

This looks like mistake number one, according to me – it assumes that the damage caused during the Panic of 2007 was all attributable to the financial crisis itself, which I think is a load of hooey.

Recessions are good things; a recession is nature’s way of telling us we’re doing things wrong. Most of the harm experienced in Canada during the Panic may be ascribed to the auto industry … hands up everybody who thinks that there were no problems in the auto industry prior to the crisis! Didn’t think so … the Panic brought things to light and forced the decision makers to address a problem … it didn’t actually cause the problem.

Financial crises are damaging because they bring a lot of problems to light all at the same time.

But for the first time, I see official acknowledgement that higher capital ratios are not automatically good:

The benefits of higher capital and liquidity standards must be weighed against their potential costs to the Canadian economy. Over time, it is expected that banks will seek to pass on the cost of higher capital and liquidity requirements through higher lending rates to borrowers. The cost of higher capital is higher lending spreads, which the Bank calculates would increase by about 14 basis points for every percentage-point increase in bank capital requirements. This figure was then used as an input to the Bank’s macroeconomic models to gauge the impact on economic output. New liquidity requirements also present costs for the Canadian economy. These requirements are estimated to add roughly an additional 14 basis points to lending spreads, and thus are equivalent in impact to an additional 1-percentage-point increase in bank capital requirements. Consequently, the cost of a 2-percentage-point increase in capital requirements, in conjunction with the new liquidity standards, should be an increase in lending spreads of about 42 basis points.

However, they show more than just a little political influence with:

In the wake of the crisis, the Bank of Canada cut its target for the overnight rate to a historic low of one-quarter of one per cent, and decades of progress in improving Canada’s fiscal position suffered a setback as fiscal support for the Canadian economy pushed federal and provincial government budgets back into substantial deficits.

The GST cut and elimination of the structural surplus had nothing to do with it, so vote Conservative!

The regulators have been busy:

To assess the potential economic implications of these reforms, the Financial Stability Board (FSB) and the BCBS conducted two international studies that assessed (i) the longer-run macroeconomic benefits and costs (the LEI report) and (ii) the shorter-term transition costs (the MAG report) associated with adopting the new standards. (footnote) The reports are: “An Assessment of the Long-Term Economic Impact of the New Regulatory Framework” (the LEI report), and “Assessing the Macroeconomic Impact of the Transition to Stronger Capital and Liquidity Requirements – Interim Report “ (the MAG report). Both are available at http://www.bis.org.

They do acknowledge a major problem:

No allowance is made for the possibility that households and firms may find cheaper alternative sources of financing in the longer run that would reduce the impact of the new rules on the economy.

If loan rates increase 42bp, I would consider that a certainty. Private mortgages will take off and I’m wondering about the potential to start a private mortgage fund myself. What’s more, this will siphon off the good business from the banks and leave them with the dregs … but there’s no allowance for it.

Specifically, assuming a starting probability of a crisis of 4.5 per cent in any given country, the LEI report found that an increase of 2 percentage points in bank capital ratios reduced the probability of a financial crisis by 2.9 percentage points, while increases in capital ratios of 4 and 6 percentage points reduced the probability of a financial crisis by 3.6 and 4 percentage points, respectively. These calculations assume a combined effect from increases in capital as well as from new liquidity rules.

When these reductions in the probability of a crisis (e.g., 2.9 percentage points for a 2-percentage-point increase in capital ratios) are multiplied by the cumulative cost of crises (63 per cent of GDP), the benefits to annual economic output are potentially large, in the order of 2 per cent of GDP

If, as I assert, the bulk of the costs experienced during financial crises are simply reflections of structural problems that are merely brought to light by the crisis, then this calculation is … er … inoperable.

For example, following the approach used by the United Kingdom Financial Services Authority (U.K. FSA) and described in Barrell et al. (2009), the annual probability of a domestic financial crisis was estimated based on several domestic factors, including the unweighted capital ratio of banks, their liquid assets as a share of total assets, and house prices expressed in real terms. This approach suggests a 1.7 per cent probability of a financial crisis occurring in Canada (implying that a financial crisis occurs, on average, every 60 years or so). This figure is substantially lower than the LEI report’s 4.5 per cent likelihood of a foreign financial crisis (approximately once every 22 years).

Since the banks nearly blew themselves in the late eighties with the MBA crisis, I guess that means we’ve got about forty years to go.

As indicated in Annex 1, most Canadian banks appear to be well placed to meet the new Liquidity Coverage Ratio requirement, since they carry a large stock of residential mortgages that could be converted at a small cost to federal government-guaranteed National Housing Act Mortgage-Backed Securities that would qualify as eligible liquid assets under the new rules.

Great! Wave a magic wand, and suddenly you get to recategorize existing assets. This part really gives me a lot of confidence, you know?

Given the current exceptionally low level of bank deposit rates and the cost of bank debt funding more generally, it is also assumed that the wider interest margins will effectively result in higher interest rates (lending spreads) on bank loans to households, firms, and other sectors of the economy. It is further assumed that the higher lending spreads will be passed along to all bank borrowers, and not just to certain subgroups, such as households or small and medium-sized businesses (SMEs), because all banks in Canada and abroad will be affected by the higher capital and liquidity requirements.

Apparently, Dr. Pangloss had a hand in this report. As noted above, increased spreads will increase the opportunity for shadow banks to move in … there will be a lot of business borrowing at prime (or prime+) who are going to find the commercial paper market (either directly or via BAs) and short-term bond issuance to be a whole lot more attractive if spreads increase 42bp.

Don’t get me wrong! I’m very happy to see that, at last, there is some official acknowledgment that increased capital standards will have a cost. But I think some of the embedded assumptions are more than just a little bit suspect and I look forward to seeing academic attacks on this paper.

A related post is Elliott: Quantifying the Effects on Lending of Increased Capital Requirements

The Federal Reserve Bank of Boston has released a discussion paper by Kristopher S. Gerardi, Christopher L. Foote, and Paul S. Willen titled Reasonable People Did Disagree: Optimism and Pessimism About the U.S. Housing Market Before the Crash:

Understanding the evolution of real-time beliefs about house price appreciation is central to understanding the U.S. housing crisis. At the peak of the recent housing cycle, both borrowers and lenders appealed to optimistic house price forecasts to justify undertaking increasingly risky loans. Many observers have argued that these rosy forecasts ignored basic theoretical and empirical evidence that pointed to a massive overvaluation of housing and thus to an inevitable and severe price decline. We revisit the boom years and show that the economics profession provided little such countervailing evidence at the time. Many economists, skeptical that a bubble existed, attempted to justify the historic run-up in housing prices based on housing fundamentals. Other economists were more uncertain, pointing to some evidence of bubble-like behavior in certain regional housing markets. Even these more skeptical economists, however, refused to take a conclusive position on whether a bubble existed. The small number of economists who argued forcefully for a bubble often did so years before the housing market peak, and thus lost a fair amount of credibility, or they make arguments fundamentally at odds with the data even ex post. For example, some economists suggested that cities where new construction was limited by zoning regulations or geography were particularly “bubble-prone,” yet the data shows that the cities with the biggest gyrations in house prices were often those at the epicenter of the new construction boom. We conclude by arguing that economic theory provides little guidance as to what should be the “correct” level of asset prices —including housing prices. Thus, while optimistic forecasts held by many market participants in 2005 turned out to be inaccurate, they were not ex ante unreasonable.

I’ll admit I was undecided about whether to highlight this paper in its own post, or simply to mention it in today’s market update … until I read the following:

It is instructive to read the logic of non-economists who looked at house price data in the same period. Paolo Pellegrini and John Paulson, whose wildly successful 2006 bet against subprime mortgages is now the stuff of Wall Street legend, made the following argument, as chronicled in Zuckerman (2009). First, they noted that house prices had deviated from trend:
…
Those facts are indisputable, but the logic that followed would have earned the two investors a zero on an undergraduate finance exam:

Ha-ha! I hope this gets quoted extensively by Fabulous Fabio and Alan Greenspan as they defend themselves against the charges that they bear personal responsibility for the Panic of 2007.

The authors don’t spare Krugman:

Krugman’s thesis seems to hinge on the idea that scarce coastal land is valuable and bubbles can only happen when assets are in short supply, but the whole point about bubbles is that the fundamentals of supply and demand do not matter. Thus, there is no reason why land in places where it is easy to build could not experience bubbles. Ex post, as we will explore at length, the places in the United States where the housing market most resembled a bubble were Phoenix and Las Vegas. According to recent research, both locations are characterized by relatively high housing-supply elasticities; unlike certain coastal areas, the two cities have an abundance of surrounding land on which to accommodate new construction.

The authors’ purpose is clear:

Ultimately, our paper argues that the academic research available in 2006 was basically inconclusive and could not convincingly support or refute any hypothesis about the future path of asset prices. Thus, investors who believed that house prices were going to fall could find evidence to support their position, while those who wanted to believe that house prices would continue to rise could not be dissuaded either. There were reasonable arguments on both sides.

One of the bubbleistas was Baker:

In addition to the divergence between rents and prices in the U.S. housing market, Baker also called attention to changes in demographic trends that could put additional downward pressure on house prices. He noted that during the 1970s and early 1980s, housing grew from about 17 percent of consumption to more than 25 percent, in large part due to increased demand for housing from the first baby boom cohorts, who were then entering adulthood. From the early 1980s to the mid-1990s, the housing share of consumption remained relatively constant, consistent with the modest demographic changes taking place in the United States at that time. In the future, Baker argued, as the baby boomers entered retirement, housing demand—and hence prices—would likely fall.

This argument has been taken up by some researchers at BIS, as discussed on August 4.

Baker also supplied contemporary arguments against the Greenspan-dunnit thesis:

As we will
discuss in more detail below, many economists pointed to low interest rates as justifying higher housing prices, but Baker was skeptical of this claim. Nominal interest rates were indeed low in the early 2000s, as the Federal Reserve had adopted a loose monetary policy to combat the effects of the 2001 recession. However, Baker pointed out that nominal rates could not explain the divergence of housing prices from fundamentals, as it is the real interest rate (the difference between the nominal rate and expected inflation) that should influence prices.

Another very interesting point is:

The evolving landscape of mortgage lending is also relevant to an ongoing debate in the literature about the direction of causality between reduced underwriting standards and higher house prices. Did lax lending standards shift out the demand curve for new homes and raise house prices, or did higher house prices reduce the chance of future loan losses, thereby encouraging lenders to relax their standards? Economists will debate this issue for some time. For our part, we simply point out that an in-depth study of lending standards would have been of little help to an economist trying to learn whether the early-to-mid 2000s increase in house prices was sustainable. If one economist argued that lax standards were fueling an unsustainable surge in house prices, another could have responded that reducing credit constraints generally brings asset prices closer to fundamental values, not farther away.

Another good point is:

If we have learned anything from this crisis, it is that large declines in house prices are always a possibility, so regulators and policymakers must take them into account when making decisions. A 30 percent fall in house prices over three years may be very difficult, if not impossible, to generate in any plausible econometric model, but a truly robust financial institution must be able to withstand one. The fact that so many professional investors as well as individual households ignored this possibility, even in 2006, suggests that we cannot allow investors to try to time market collapses.

All in all, most interesting and well balanced. Related posts on PrefBlog include:

• Hull & White on AAA Tranches of Subprime
• Bernanke: Monetary Policy and the Housing Bubble
• Subprime! Problems forseeable in 2005?

As for me … I’ve always disclaimed any ability or interest in forecasting macroeconomic trends. But what I have said is … bad investments will hurt you. Concentration will kill you.

The Panic of 2007 wasn’t caused by Merrill Lynch et al. buying sub-prime paper. It was caused by the fact that they levered it up big time.