Weekend long reads (Dec. 5, 2014)

Readers will notice that there hasn’t been much activity here since Thanksgiving. My absence is partly due to traveling I’ve had to do, being engrossed in my new book (The Power Broker by Robert Caro), and other academic obligations, which will continue next week. Nonetheless, I have provided some long reads here as they seem to one of the more popular types of posts. I hope to have the third part of the series on sectoral investment patterns up by the end of next week.

Fracking tantrums

Banking

Research and academics

Should Lehman Brothers have been allowed to fail?

On the question, briefly

I was recently asked the question: “Should the investment bank Lehman Brothers have been allowed to fail in the summer of 2008”? I’d like to sketch out a couple of points about the question and its answer, mainly to make myself feel better after realizing I did not have an answer ready and responded so inarticulately.

Background

For a more thorough introduction to Lehman Brothers, I suggest scanning this post at Investopedia (http://bit.ly/1mJP8nM) and the Wikipedia entry entitled ‘Bankruptcy of Lehman Brothers’ (http://bit.ly/ZXfoBe). A search for ‘Lehman Brothers’ on Amazon reveals a number of books about the firm, none of which I have read. Here are some key points.

  • Lehman Brothers was an American, New York-headquartered investment bank–fourth largest in the States–with global operations. It had been part of American Express until that company decided to divest its banking and brokerage operations in 1994. The new entity, Lehman Brothers Holdings, Inc. was listed on the New York Stock Exchange. The company operated in three segments: capital markets; investment banking; and investment management.
  • In 2003 and 2004, the firm acquired mortgage lenders that were engaged in high-risk practices, namely subprime lending. The source of growing revenues in its capital markets division was securitization techniques, and revenues in this division surpassed other areas in the company shortly after acquisition. Record profits were reported from 2005 to 2007. Incidentally, rapid growth in profits is a tell-tale sign of an accounting control fraud.
  • Defaults on subprime mortgages began to increase over late 2006-early 2007. By August 2007, Lehman stock dropped substantially following the failure of two hedge funds associated with investment bank Bear Stearns. Lehman began shuttering its mortgage businesses, but continued to underwrite mortgage-backed securities. Global equities and fixed-income markets brushed off the hesitations earlier that year that profitability would suffer as mortgage and housing markets registered rising defaults.
  • Bear Stearns collapsed in March of 2008, and Lehman’s stock price dropped by almost 50 percent. Lehman was able to raise capital through preferred stock, however. In June, the company announced a loss of $2.8 billion but continued to raise capital from investors. It further announced it had increased liquidity, decreased assets, lowered exposure to mortgages, and decreased leverage. The stock price continued to drop over the summer and investors spurned management.
  • In September, the collapse in talks between Lehman and Korea Development Bank precipitated a new drop in the stock price as investors (hedge funds and short-term creditors) began to withdraw funds and limit exposure. The stock continued to drop as its cash reserves dwindled, losses and write-downs were reported, and restructuring programs were announced. A final attempt at a rescue by Barclays and Bank of America failed over the weekend of September 13 and on that Monday, the company declared bankruptcy.
  • Bear Stearns was in a similar position to Lehman Brothers, in that it was not a chartered institution, was publicly-listed, was an investment bank, and faced financial distress from its risky positions and unsound management practices. It was, however, the recipient of an emergency Federal Reserve loan via JP Morgan, earlier in the summer. The reasons for this, contrasted against the lack of such a loan to Lehman, are discussed below.

Is the question relevant?

My first comment is that the phrasing of the question “should Lehman have been allowed to fail” is actually quite misleading. A more pertinent question is “Could Lehman have been saved?” The answer, I think, is no. Recall that the rescue of Long-Term Capital Management in 1998 was orchestrated by the Federal Reserve of New York, but involved no public funds. That is, the Fed compelled creditors of Long-Term to recapitalize the firm. There were several attempts at a private sector recapitalization of Lehman Brothers; there was simply no appetite by its creditors or competitors to intervene.

Why is it the case that after the inability to recapitalize privately there was no recourse to a public recapitalization or any other preventative action by a government agency? Just as was the situation with Long-Term, Lehman was not a chartered bank. It’s primary regulator was the Securities Exchange Commission (SEC); the FDIC and (New York) Federal Reserve were not its supervisors. The SEC has no powers like the FDIC or Fed, and can really only intervene through review of a company’s risk management and investor protection policies. It can also levy fines on its supervised entities. In other words, it probably would have been illegal for the FDIC or the Fed to have used public funds to support Lehman Brothers. Furthermore, the SEC and the New York Fed have a long and combative history. Again, even though the Fed could not have provided any kind of balance sheet support, we should at least recognize that these turf wars prevented a lot of meaningful interagency communication.

In the absence of any consensus by private companies to support the bank, there was very little the state could do within the parameters of existing banking and macroprudential law to prevent the bankruptcy of Lehman. Specifically, with no chartered bank, the FDIC could not intervene by seizing the bank and its assets. The US Treasury could not have organized a direct loan nor could it have offered funds under various emergency programs related to the preserving the integrity of the dollar. And the SEC has no such powers in any event.

The exception is Section 13(3) of the Federal Reserve Act, which authorizes the Federal Reserve Board to permit a Federal Reserve Bank to extend loans under exigent and unusual circumstances. So essentially the question is really why Lehman Brothers was not offered an emergency loan as was Bear Stearns. What explains this discrepancy? In the words of Fed Chairman Ben Bernanke:

A public-sector solution for Lehman proved infeasible, as the firm could not post sufficient collateral to provide reasonable assurance that a loan from the Federal Reserve would be repaid, and the Treasury did not have the authority to absorb billions of dollars of expected losses to facilitate Lehman’s acquisition by another firm. Consequently, little could be done except to attempt to ameliorate the effects of Lehman’s failure on the financial system (underline mine).

The underlined portion is actually very important. Pursuant to Section 13(3) of the above stated Act, the Fed cannot extend a loan with the knowledge that the loan cannot be repaid. The perception on Wall Street prior to Lehman’s bankruptcy was that it was insolvent; it could not repay any loans extended to it. Certainly, this does not explain why the Fed did not extend such a loan many months before, but then hindsight is twenty-twenty, and there are limits to what these regulators know (especially in light of the turf wars just mentioned). So, the state was effectively unable to save Lehman Brothers.

So what?

The next pertinent question is, what does it matter that Lehman Brothers failed? Many would stipulate that its bankruptcy was the triggering mechanism of the banking panic that fall. I suggest this is wrong. There are several relevant points here.

First, housing markets began to buckle in late 2006 and early 2007. This manifested in liquidity problems at a host of non-bank financial institutions, including two hedge funds associated with Brazilian firm BNP Paribas in 2007 and Bear Stearns as mentioned. Saving Lehman would not have prevented housing markets from tanking or prevented the liquidity problems associated with the failure of mortgage markets and products. The Lehman bankruptcy was only one event in a long series of events that dried up credit and lending. Interbank lending in particular had already seized up.

Second, it is arguable that US financial markets were already in a state of panic before the weekend that Lehman failed. Again, note the experience of Bear Stearns. There are other examples: Countrywide Financial experienced a run in August 2007 before being acquired by Bank of America; IndyMac Bank suffered a run in July 2008 after Senator Charles Schumer stated to the media that he believed the bank was unstable; Washington Mutual witnessed massive withdrawals by its retail customers beginning on the same day as Lehman announced its bankruptcy. The financial panic, not only in the ‘shadow’ financial sector as well as in retail, therefore, was already underway when Lehman failed.

And finally, the stock markets “crashed” beginning on September 16, 2008, the day after Lehman declared bankruptcy. Indeed, during several trading sessions in September and October of 2008, US stock markets lost two to five percent of their value. However, again, equity markets had begun to decline beginning in 2007. If Lehman Brothers had not failed, then stock markets likely would have responded to similar triggering mechanisms, of which there were many.

These points support the conclusion that at a certain stage in the financial evolution of the US economy between 2003 and 2006, a financial crisis became inevitable. This was due to the construction of balance sheets, the dependencies erected between sectors of the economy, the rampant fraud and corruption in markets, etc. Lehman stands out principally due to, I suspect, the animus held by many on Wall Street for state regulation and action that doesn’t suit their wishes. The Lehman Brothers bankruptcy is partly an instance of scapegoating–‘the government failed to act’ narrative–and a failure of imagination, specifically the failure to grasp the roots of financial crises as endogenous and nonrandom.

The loan to deposit ratio and the trouble with US finance

 The three problems in US finance revisited

In an earlier post, I suggested there were three core problems with US finance: tendency towards dysfunction; uneven cost and access to financial services; and, waste, redundancy, and fraud. I was just browsing through blogs and saw a post at zerohedge (http://bit.ly/1so7Gd9). It is entitled ‘All that is broken with the US financial system in one chart’ and shows the loan to deposit ratio at JP Morgan from 2008 to now. The question I’m asking myself is: does JPM’s loan to deposit ratio really tell us everything wrong with US finance? Does it reveal the US tendency to undermine profitability and stability by way of incentives to misallocate credit? Does it reveal the uneven ability for consumers to access credit on good terms? Does it reveal the wastefulness of the US system, the high incidence of consumer fraud, and the high incidence of white-collar fraud?

The first thing I note is that I don’t think the 2008 to 2014 period is satisfactory for answering these questions. Furthermore, there is no comparison to other large banks, even though JPM on its own has market power and can be considered a special case worthy of analysis. I don’t post comments on that blog, but I did notice in the comments section someone asking if there was a chart for the whole financial system. The implicit criticism, I think, was that JPM is not indicative of the rest of the banking system, even though it is the largest bank in terms of deposits, or at least in the top four. I think a better starting point is asking whether or not the loan to deposit ratio can capture the problems in US finance, and then proceeding with individual banks to understand how the ratio is constructed organizationally and geographically.

I’ve attached below a chart created using FRED (http://research.stlouisfed.org/fred2/), showing the ratio of loans & leases to deposits at US commercial banks. It begins in 1973 and continues monthly to the start of October 2014.

Loan to deposit ratio, US commercial banks

It seems to me this is another important institutional relationship in US finance. Some notable features include: (1) the ratio currently stands at its lowest level since the late 1970s; (2) it peaked during the dot.com bubble, and even the mortgage lending boom could not raise the ratio to its previous peak (which suggests some interesting limits to leverage); (3) it increased rapidly starting in the mid-1990s economic boom; and, (4) it had been in decline from the panic in 2008 until earlier this year.

There are a number of different behaviors that affect the ratio. First is clearly the demand for credit, which is structured by access and cost. Demand, by this token, is affected by the organization of credit allocation, that is the specific institutions responsible for providing it (too-big-to-fail banks, savings associations, credit unions, payday lending). Second is the effect of interest rates, which affect willingness to take on risk by banks. Third is capital and other regulatory requirements.

So, does this institutional relationship reveal the problems in US finance? This question assumes that the three-problems are valid; they may not be, and this is a good exercise to work that issue out.

1. Tendency towards dysfunction

The ratio tends to drop around US recessions, but it is interesting to note that it did not drop as much during the S&L debacle. Leverage actually appears to have increased from 1985 onwards, even through 1987 when there was a sharp drop in profitability of US banks. Even during a banking crisis, leverage continued to grow. It did, however, respond dramatically to the latest recession and banking crisis, however it is hard to separate out those effects. Did leverage decline after 2008 because of demand or because of supply side factors? I don’t know. At the least, the ratio might anticipate collapses, in that periods of expansion in leverage are followed by contractions, which is to be expected based on periods of credit expansion.

2. Cost and access is uneven

In order for us to determine whether or not the ratio can reveal the problem of access, it would probably be best to disaggregate the ratio between types of organizations. First would be sector (commercial banks, savings associations, credit unions, non-regulated intermediaries); second would be size of organization; and third would be the region of operations. It would be great if there were data on the average ratio by banking markets served (which could then be related to other demographic features of those markets, such as income, population, etc), but that would be quite an undertaking. At least in general terms, the ratio suggests that there has been a significant pullback in leverage; as we cannot quantify the factors that contributed to that (it could be demand, regulation, and supply), it is difficult to know whether or not the pullback reflects lower willingness by consumers to take on debt or greater risk aversion by banks. The post at zerohedge argues that the flatlining in amount of loans outstanding indicates a conscious decision by the bank not to extend loans; I think we would need more data on the volume of loan applications as well as the rate at which applications are accepted. The Federal Reserve has survey data from banks, which speaks in a way to sentiment, but hard data would be more useful.

Another important point that comes to mind, and which returns to the issue of access and organization type, is that using commercial banks to determine lending might actually be misleading. As I’ve posted before, regulated banks are actually performing less lending. The ‘shadow’ financial sector has been encroaching on traditional lending markets for decades. It may be the case that the growth of shadow activities in financial markets has lowered the share of commercial banks in lending. To develop this line of inquiry further, one would need to explain that this pattern accelerated after 2008.

3. Waste and fraud

On the whole, the ratio really seems to be an indicator of system-wide leverage. It does not seem capable of tackling issues of consumer fraud, white-collar crime, etc.

Summary

I am weary of these reductionist articles that pop up on so many blogs and online news sites. The problems in the US financial system cannot be boiled down to any one factor. The sector is too heterogeneous for there to be even one set of factors. Clearly, there are more than three problems as well, but I’ve tried to make it so that other “sub-problems” can be filed under those categories. Again, the endeavor to determine just exactly what is wrong with US finance has to take place on a lot of levels–regional, sectoral, historical, institutional, etc. The key may be to generate case studies that are regionally and institutionally specific and see what kinds of patterns and differences can be identified. That said, I’m glad to add the ratio to the list of relevant institutions of finance.

Too-big-to-fail profitability

This is a short post, comparing basic financial information of the too-big-to-fail banks (these are: Bank of New York Mellon, Bank of America, Citigroup, Goldman Sachs, JP Morgan, Morgan Stanley, State Street, and Wells Fargo). Of these organizations, three are not headquartered in New York (BOA in Charlotte; State Street in Boston; WFC in San Francisco/Minneapolis).

The information I have included is the change in stock price over the 2008 crisis and price performance since the trough up to now; return on assets; a liquidity ratio (from the balance sheet, cash as a ratio to total current liabilities); and, profit as a share of total revenue. All this information can be collected from the balance sheets and income statements of the banks, which I retrieved using Google. The figure below contains the information.

Stock price change ROA Liquidity (Cash/Current liabilities) Profit/Revenue
2007-2009 2009-2014 June2014 June2013 avg 2013 avg 2012 avg 2013
BNYM -63% 58% 0.74% 0.56% 0.5 0.49 0.137
BOA -94% 129% 0.37% 0.30% 0.39 0.4 0.112
Citi -95% 42% 0.52% 0.51% 0.35 0.32 0.180
Goldman -77% 71% 0.87% 0.91% 0.29 0.285 0.235
JPM -70% 267% 0.66% 1.03% 0.35 0.275 0.188
MS -83% 70% 0.53% 0.19% 0.3 0.51 0.090
State Street -75% 74% 0.85% 1.02% 0.88 0.93 0.216
WFC -77% 563% 1.50% 1.48% 0.44 0.42 0.265
Average -80% 181% 0.82% 0.86% 0.44 0.46 0.196

There are a few points I see as especially relevant. First, almost none of the banks have restored the price of their stock to levels seen around 2007, the stock market peak. Not all of the banks, it should be said, were trading at historical peaks in 2007; nonetheless, they all responded to the crash in 2008. Bank of New York Mellon, State Street, JP Morgan, and Wells Fargo had below-average declines, but only JP Morgan and Wells had recovered their losses. The worst-performer has been Citigroup, which comes as no surprise seeing as it is easily the worst-run organization on the planet.

Second, the range in ROA is quite large. I’ve spent a lot of time looking at ROA at lots of other banks of various sizes and specialization; a previous post here put US total average bank profitability currently at around 1%. Wells is the only company reporting an ROA greater than the US average (in 2014; many more were closer to the average about a year ago). I’m not sure what the patterns are–perhaps there are regional trends, but I suspect they have much more to do with target deposit markets and product lines, etc–but this variability in ROA encourages further analysis.

Third, it is illuminating to see the liquidity ratio comparison if only because it introduces the challenge faced by regulators and supervisors on macro-prudential matters. Current liabilities consist of accounts payable, short-term/current long-term debt, and ‘other current liabilities.’ Each of the banks, in turn, have different concentrations in each of these areas. The amount of cash (and current liabilities) also differs substantially between companies. Obviously, one of the more quiet fights by regulators over the last few years has been to raise capital ratios (one way to ensure that costs of crisis resolution fall as much on the private sector as possible). Some of these businesses simply operate in riskier markets; this may explain why there is such variation in liquidity ratios. I wonder how this unevenness is looked at by the regulators themselves?

Finally, the profit as a share of revenues for all too-big-to-fail banks is close to 20 cents on the dollar. I’m not sure what this number is for all Fortune 500 companies, but I remember hearing an interview with the CEO of ExxonMobil some years ago where he said that Exxon’s profit per revenue was 10 cents on the dollar, and that this put Exxon ‘right in the middle’ of the Fortune 500. If you look at the most recent Fortune 500 list, the top four companies (oil majors and conglomerates) are all less than 10%, but then Apple is at 21.6%. What this all suggests to me is that financial conglomerates enjoy profit margins that are much higher than capital goods producers and energy companies, but then there remains quite a bit of difference. Again, I suspect this has to do with product and market specializations but also company structure.

It is easy to clump the too-big-to-fail firms together, but in fact they serve quite different markets. When we devise strategies and policies to control them, we need to take into account this heterogeneity and plan accordingly.

Bank profitability compared

I’m trying to find quantitative indicators of bank profitability and activity that might be compiled into an index of macro-prudential stability at sub-national, regional scales. Ideally, such an index or set of indicators would be able to identify thresholds around which sector profitability is endangered or to identify mid-term growth potential for the area at large.

At the national-level, data for aggregate bank profitability is available. The World Bank collects data ostensibly from national agencies on the return on bank assets. The graph below was created using the FRED (http://research.stlouisfed.org/fred2/) function of the St. Louis Federal Reserve. It compares the return on bank assets for Australia, Canada, the United Kingdom, and the United States.

bank-return-assets-four

The dataset appears to go back only to 1998 and is annual, which is a shame as quarterly fluctuations can be informative. It is interesting that Canada has demonstrated a relatively consistent upward trend since 2000, while the United States had a steady level of profitability from 2000 to 2006. Australia and the United Kingdom are characterized by rather large swings.

The Federal Financial Institutions Examination Council (FFIEC) is responsible for collating similar data for US banks, which is no small task as there are several thousands of banks in the US. That data is collected quarterly, and extends back to the early 1980s. The graph representing US bank profitability (also return on assets) is available below.

us-bank-return-ffiec

This graph indicates that American banks enjoyed a long period of stable profits. The S&L crisis is clearly evident in the late 1980s, although it seems that despite a brief recovery, for several years after the immediate crisis years, profits remained depressed. It was not until around 1991/2 that profitability shifted to a higher plateau. Also notable is that profitability increased before the economic boom from the information technology revolution, which did not really begin in earnest until 1994/5. An additional important feature is that, although profitability had returned to positive levels after the 2008 crisis, it remains below its average levels during the preceding twenty-year period.

While profitability is a good indicator of stability, it has its limits. From my perspective, I wonder how possible it would be to disaggregate profitability between regions and types of banks (for instance, depending on their loan portfolio concentrations [commercial, industrial, residential, etc]). That is, it would be very enlightening to know how banks in different parts of the country are contributing to or perhaps retarding economic growth in general through their impact on the cost and type of loans, raising deposits, and their management of non-performing loans. In particular, it would be interesting to examine the extent to which bank activity in a given area relies on deposit-taking or recourse to capital markets.

In the meantime, I’m digging around for comprehensive datasets that may have this information. Otherwise, it would be seem to be the case that one would have to aggregate this data from bank call reports, which given the number of banks and the time scale necessary for meaningful analysis, is a truly onerous task.

American financial dysfunction

Three problems in US finance

US financial dysfunction manifests itself outwardly in at least three characteristics. First, the system as a whole displays a remarkable tendency towards catastrophic failure, in terms of severe contractions in profitability, widespread insolvency of organizations, and periodic illiquidity. Second, access to and cost of credit is uneven, across regions and between social groups. And third, there is quite a bit of redundancy, waste, and fraud. The fraud exists at two levels. First is the problem of white-collar crime, where executives in banks and other financial organizations use firm competencies and knowledge as weapons for their personal aggrandizement; and the second refers to the kind of fraud that most consumers experience (theft of personal financial information) as well as attacks on the integrity of bank electronic and security systems.

My reason for framing US financial dysfunction in this way is that I don’t see a coherent reform program that tackles the main problems of US finance. Often we hear about ‘breaking up the banks’, or the problem of cognitive capture of the regulators, the lack of transparency in securities markets, etc. These are problems that we fixated on following the 2008 crisis, but it is worth mentioning that there were efforts at financial reform during the 2000s (Sarbanes-Oxley, the bankruptcy reforms in 2005, attempts to regulate credit rating agencies) that didn’t produce a more stable, secure, or efficient financial system. The troubles ran, and still run, deeper. My objection to the current proposals, besides the fact that they rarely appear together in any kind of blueprint that weighs how they will be implemented, the ramifications for doing so, and economic rationale, is that they do not do anything for the long-standing problems in US finance: tendency to crisis, access/cost of credit, and fraud.

Essentially, these three problems, being unique to the US, reflect the outcome of the ‘Game of Bank Bargains’ as it has played out in the US (based on the analysis by Charles Calomiris and Stephen Haber in their book Fragile by Design). Calomiris and Haber discuss the problems of tendency to crisis and cost of credit, but do not focus much on the fraud angle. They also do not spend much time on two elements that I find particularly important: the types of organizations that are permitted to allocate credit (public sources, private institutions, and their specific institutional profiles) and the types of organizations that are set up to supervise the allocation of credit (centralized/decentralized agency; independent agency, or democratically accountable; financed by appropriations, or not; and the rest of its institutional profile).

The point of this post is to gather some evidence for the first of these problems, and to determine how the US fits in amongst its cohort of rich countries. In subsequent posts, I will gather the evidence for the other two points, and then be at a point where I bring all the evidence together to determine whether or not there is a solid empirical basis for the following paradox: how is that a rich and (formerly?) prosperous country like the United States can feature such a destructive element at the core of its economy?

Hopefully as I analyze the data, the argument will evolve and become more nuanced, if not thrown out altogether. Nonetheless, I think it is a good hypothesis and I think it already has quite a bit of support. I believe the answer to why this paradox exists has to do with the characteristics just mentioned, but, more specifically, I think it is important to highlight how the structure of finance systematically displaces the costs of its excesses and failures onto other parts of the economy and society. This conclusion, of course, assumes that the paradox exists.

The historical tendency to crisis

We really should not be surprised that there is this highly destructive element at the center of the economy, because capitalism is propelled by creative destruction. Yet the US stands out because its level of financial dysfunction is so much greater than peer countries. One way of showing this is to compare the US to a group of countries that share many of its institutional and economic characteristics. I’ve selected Australia, Canada, and the United Kingdom for this brief example.

The first step is to establish that the US is more prone to crisis than these countries. Using data from Carmen Reinhart and Ken Rogoff (http://www.reinhartandrogoff.com/data/browse-by-topic/topics/7/), I tally the number of years in which there was a stock market crash or banking crisis in the four countries between 1945 and 2010. The figures are collected below.

Years between 1945 and 2010 spent in a
Stock market crash Banking crisis
Australia 10 4
Canada 5 3
UK 11 9
USA 21 12

In their book This Time is Different, Reinhart and Rogoff define these crises are as such. A stock market crash is where real equity prices decline by 25 percent. A banking crisis features bank runs leading to government intervention, or government intervention into the banking sector (closure, merger, takeover by public sector of a single or group of financial institutions). Canada is by far the most stable, while the USA has spent the most number of years in a crisis (almost a third of the post-war period has witnessed dramatic crashes in equity markets!). The UK and Australia are about the same for number of years of stock market crashes, but the UK is closer to the USA for number of years in a banking crisis.

Obviously, the years are clustered in time because a crisis often spans many years. For example, in terms of episodes, the USA has much longer banking crises than the UK. In the US, there was the S&L crisis, which Reinhart and Rogoff record from 1984 to 1991, and the recent crisis from 2007 to 2010 (when the dataset ends). The UK, by contrast, has a three-year crisis beginning in 1974, and shorter crises in 1984, 1991, and 1995, and then the recent one from 2007 to 2009. For Australia, there was a four-year crisis beginning in 1989, and in Canada, a three-year crisis beginning in 1983.

Much of the difference might reflect the fact that the US has had historically a very large banking population, as a consequence of the New Deal reforms (intra- and inter-state banking restrictions). In the US, once a banking crisis begins, it can affect so many more organizations before burning out. I will return to this point in another post.

Another dataset on banking crises has been compiled by economists working at the IMF. This dataset includes only “systemic” banking crises (available here: https://www.imf.org/external/pubs/cat/longres.aspx?sk=26015.0). These crises include the following elements: defaults of institutions in the corporate and financial sectors of a country; difficulty of institutions in these sectors in the timely repayment of contracts; an increase in non-performing loans; illiquidity; depressed asset prices following the crisis; rise in real interest rates; and, potentially, capital flow reversal. Such a definition is quite different from the banking crises in the Reinhart and Rogoff dataset, namely in that it attempts to include quantitative indicators.

Using that information, the IMF finds no systemic crises in Australia or Canada, and only in 2007 was the UK in a systemic crisis. In the USA, there were two crises beginning in 1988 and 2007. Even after adopting a stricter definition of a financial crisis, the USA surpasses its immediate peers in financial dysfunction.

Alternative indicators

Banking crises are more difficult to pinpoint historically because bank assets do not reveal conditions of crisis the same way that equity or real estate prices reveal a crisis in asset markets. This feature partly explains the loose definitions above and the reliance on qualitative indicators, such as government intervention. One criticism of using government intervention as an indicator of crisis is that it can become a self-fulfilling prophecy, while another is that there remains the question as to what level of intervention counts as a crisis. It may also be the case that individual countries or groups or types of countries (democratic, autocratic, young/old countries, English common law system, industrial-based) experience different types as well as different manifestations of crises. As eluded earlier, a banking crisis in the USA looks very different from one in Canada given the large number of unit-banks, the highly compartmentalized regulatory system, and the high level of capital mobility. My preference is to situate “crisis” in the context of these country-specific factors. Nonetheless, there is some value in comparative work.

Institutional change in lending

Who lends?

The declining share of deposit money banks in total credit lending indicates increasing diversification and specialization of capital markets. Organizations that are not deposit money banks include credit unions, money market funds, pension funds, insurance agencies, brokerages and other securities firms, financing companies, venture capital firms, hedge funds, and, at the lowest-tier of the financial structure, payday and other short-term, small amount lending.

Another way of interpreting the share of deposit money banks in total lending is as a share of lending that is not insured or supervised by the government. This is a bit misleading because there are public and private forms of insurance available to these organizations. Credit unions are insured by the National Credit Union Administration (NCUA, the equivalent of the FDIC), while most brokerage houses are insured by the Securities Investor Protection Corporation (SIPC), which has many of the same features as the FDIC. These programs insure depositors against losses; they do not include safety and soundness oversight. As such, the non-deposit money bank share is a reasonably good approximation of the amount of activity taking place outside the purview of public supervision.

The share of deposit money banks in lending is an important institutional relationship for a national financial system. Data for this is collected in the US by the Federal Reserve and other agencies, and the Bank of International Settlements has put together a national cross-sectional, time-series dataset that can be found online (http://www.bis.org/statistics/credtopriv.htm). In the United States, this share rose from the 1940s until the 1970s, peaking below 60% before decreasing from the mid-1970s until today. It dropped substantially during the S&L crisis, as many hundreds of savings institutions but also commercial banks failed. The shares for the US, Australia, and Canada are presented in the figure below.

us-aus-can

I include Australia and Canada for the reason that the three countries share similar geographical, historical, and institutional characteristics. They are large, demographically-heterogenous, continental-sized economies that are former British colonies, and so have legal systems based on English common law. Natural resource extraction and trade are essential components of their economic growth. Their experiences with financial crises, however, are somewhat different. For instance, the US has seen two banking crises in the last thirty years (S&L and 2008), while Australia experienced a banking crisis in the early 1990s (when a couple of government-owned savings institutions failed and several other institutions), but was largely spared from the 2008 crisis. Canada has seen periodic but isolated failures of banking organizations, but was also spared from the 2008 crisis. In effect, only the US experienced a ‘systemic crisis’ during the last three decades, which was in 2008. This disparity is even more remarkable in light of the fact that in all three countries, property markets underwent an incredible boom leading up to the 2008 crash.

A institutional model of financial crises

The issue of who lends is vital when considering why some countries are more prone to financial dysfunction than others. Charles Calomiris and Stephen Haber have argued in their book Fragile by Design that the frequency with which a country experiences financial crises is a result of how its fundamental political institutions distribute power between competing groups in society. In the US, historically, populists have exerted a level of influence over the structure of banking markets due to the alliances they have formed with local bankers. In Canada, however, the federal government has insulated the banking sector from populist influence and created a more stable framework that allows its few, very large banks to control much of the lending markets while also providing abundant and cheap credit. The US, by contrast, ranks comparatively low among high-income countries in providing affordable credit (witness the lack of affordable and government-insured borrowing options for low-income Americans in poor neighborhoods and cities).

The conceptual model–the Game of Bank Bargains–that Calomiris and Haber offer is a good counterargument to explanations of financial crises that invoke ‘speculative euphoria’ or ‘banker folly’ or other aspects of human behavior. Their argument has its limits, though. For example, they spend a great deal of time repeating the conservative bugaboo about the 2008 crisis that Frannie Mae, Freddie Mac, and the Community Reinvestment Act of 1970 distorted lending incentives and caused a deterioration of underwriting standards during the housing bubble. They also suggest that community activists teamed up with mega-bank interests to push through the financial liberalization reform program of the 1990s in exchange for greater lending to urban minorities. I think in doing so they alienated a number of readers who might have been inclined to use their model but were put off by the polarizing and discredited notion that the victims of the 2008 crisis–urban minorities–were responsible for the crisis.

This is obviously a shame, because I think it is crucial that we have a more informed conversation about how financial markets and specific banking organizations are used by the state to advance the state’s interests as well as to advance parochial or regional interests. Nonetheless, it is possible to use their model of interest group collusion to tell a different story about the 2008 crisis and why credit is allocated the way it is. The central process in that story is the transformation of legal foundation of the banking sector and the change in contractual relations between households, the financial sector, and the state. I’m going to outline this story and emphasize its geographical component.

A brief application/reinterpretation of the Game of Bank Bargains

At the end of the S&L crisis, the surviving, well-capitalized financial organizations were able to take advantage of (1) a buyer’s market in distressed financial assets, and (2) an appetite for regulatory reform at the national level, aided by an accommodative Federal Reserve Board. The S&L crisis affected mainly (obviously) savings institutions in the south and southwest states. The surviving banks were money center and investment banks in New York and other northeast financial centers, as well as the so-called super-regionals. For a more in-depth review of this period, I suggest you read a working paper by Gary Dymski entitled “Genie out of the Bottle: The Evolution of Too-Big-to-Fail Policy and Banking Strategy in the US.”

Suffice it to say that the type of organizations that were poised to reap the benefits of the post-S&L environment very much shaped the nature of the legal reforms of the 1990s. That is, the reforms included repealing intra- and inter-state branching restrictions, removing separations between investment and retail activities, and repealing size limitations. Less talked about consequences of these reforms include that the role of the Federal Reserve took on greater importance. For example, under the Riegle-Neal Act of 1994, the Fed was charged with approving mergers-and-acquisition (M&A) activity, while the organizational model of choice for accomplishing M&A activity was the bank holding company. It so happens that the bank holding company is supervised by the Federal Reserve.

An even less talked about event during this period, which likely had a strong effect on the nature of bank strategy, was the experience with Long-Term Capital Management. Without getting too deep into the details, the collapse of Long-Term in 1998 was marked by an incredible intervention by the Federal Reserve, which organized a private sector bailout by leaning on Long-Term’s creditors. Some of the lessons of that experience were: the Fed was willing to drastically intervene to prevent the failure of what it considered a systemically important institution; the use of complex instruments such as derivatives would continue as a risk management mechanism; and, a highly connected firm could be successfully resolved in the event of a crisis. I suspect the experience of Long-Term was on the minds of decision-makers at the Fed and elsewhere in the upper echelons during 2008 (it had only been a decade, after all).

So, who is at the table of the Game of Bank Bargains in the 1990s? The table is populated by New York investment and money center banks, super-regionals, the Federal Reserve, and various senators and administration officials representing those organizations. I think it is fair to say that there was nothing inevitable about this particular group advocating for the particular package of liberalization reforms. It seems quite opportunistic, in fact. The group’s identity reflect the survivors of a crisis in the housing finance and commercial property sectors in the 1980s and their strategic considerations.

Was there a viable alternative to their view of the ideal financial landscape? Not really. The savings association industry was in a shambles, and community, consumer, and housing advocates had made major gains in the reforms passed during and immediately after the S&L crisis (notably by imposing as much of the costs of the S&L resolutions on the industry itself). Their continued efforts were stymied by the George H.W. Bush administration. There doesn’t seem to be much in the way of any advocacy for a greater role of public options in banking, such as state banks (in the style of the Bank of North Dakota) or perhaps a postal savings system during this time.

So the legal foundation of the financial sector was radically altered between the early 1990s up to the early 2000s. The preferred organization for consolidation was the bank or financial holding company, the preferred regulator was the Federal Reserve, and the preferred solution to the bank profitability crisis witnessed in the 1980s was the removal of geographic barriers in banking markets. Another implicit element of the Game of Bank Bargains in this instance was the too-big-to-fail issue. The bailouts during the S&L crisis demonstrated the possibility that market mechanisms could be used to resolve failed assets, while the government emphasized its willingness to intervene directly into markets to prevent disruption. The effect of these reforms was actually, as shown in the figure above, to narrow the government’s zone of control over lending–new intermediaries popped up in the interstices between the big banks, which themselves were executing more trades and lending outside of their chartered, depository units.

The Bank Bargain of the 1990s was truly a new contract between a quasi-governmental agency (the Federal Reserve) and a collection of large institutions (many of which were already embedded within international financial markets) for control over the credit allocation process. It was guided by an ethos of liberalization that had been brewing before the S&L crisis, and in fact could arguably be seen as having instigated the crisis or at least made it worse (for instance, by deregulating interest rates under DIDMCA of 1990 and Garn-St. Germain Act of 1982).

Bottom line

One of the lessons for institutional change in lending, which is how I started this post, is that the choice of organization actually matters very much. Different types of organizations face different strategic considerations and observe specific locational distributions. Additionally, types of organizations fall more or less within the state’s influence. The Bank Bargain of the 1990s sought, given the ideological climate of the time, to limit the zone of control. One of the points I picked up on reading Calomiris and Haber’s account of the banking stability of Canada is that the state endeavored to maintain or increase the domain of lending covered by its preferred organizations (its large chartered banks) as new technologies and innovations arose. The example of REITs in the US (which precipitated a crisis during the 1970s, see Hyman Minsky’s Stabilizing an Unstable Economy) shows that while REITs were eventually legitimated, they remained an industry operationally separate from the chartered and regulated system.

These are only some preliminary thoughts, but I think they provide good support for a more thorough comparative analysis of the politics and institutions of credit allocation. The US, Australia, and Canada, I suggest, are excellent candidates for such an analysis. The conceptual model by Calomiris and Haber could be reapplied by adding the spatial perspective–the location of organizations, their structure, strategies, the competitive environment–and also by reframing the issue as the extent to which a country manages (and wants to manage) the quality of its lending.