On simplifying without simplification

By skepticlawyer

[Every now and then, the law throws up something really fascinating. This week’s exercise in fascination was instructions to prepare a briefing note on the causes and consequences of the global financial crisis, directed at yours truly.

What made these instructions so difficult to follow is that while the causes and consequences of the GFC are beginning to come clear (with, of course, the benefit of 20/20 vision in hindsight), explaining them is anything but an exercise in clarity, especially in just half-a-dozen pages (which is pretty much all one gets in a briefing note).

I set myself a challenge, then: to simplify without simplification, especially when it comes to the financial products everyone was selling like billy-oh and that no-one seemed to understand (buyers and sellers both).

In what follows, names have been changed to protect the innocent (and the guilty!), and a couple of bits have been excised. I have added a couple of links, but only to concepts with which non-lawyers may be unfamiliar (the intended audience is other solicitors). The document is meant, by and large, to stand alone. Does it, I wonder? 



In 1921, economist Frank Knight distinguished risk from uncertainty on the basis that the former was susceptible of measurement, while the latter was not. This famous distinction captures the essence of the 2007-8 global financial crisis, which manifested as a liquidity contraction. In short, banks and other financial institutions first stopped lending to each other and then stopped lending to business more generally on the basis that they were no longer able to price risk. ‘Risk’ had become ‘uncertainty’: no one who mattered knew what anything that mattered was worth any more.

In this briefing, I will first outline the significant aspects of the global financial crisis and consider some of its causes, followed by a discussion of industry responses. Next, I will address the vexed question of whether we—in the UK at least—are still in ‘crisis’. Finally, I turn to issues that are pertinent to our client bank. Throughout, I will do two things: first, keep Knight’s distinction between risk and uncertainty foremost in my analysis; and secondly, distinguish—particularly when sharing the blame around—between government behaviour and industry behaviour. Where the two intersected—as in the US subprime mortgage market—I explore why that might have been the case.

The Global Financial Crisis: a recap

The immediate trigger for the GFC differed between jurisdictions: in the US, the ‘housing bubble’ burst. In Britain, banks were unable to borrow from each other in short-term money markets. It is simplest to consider the US first, as the integrated nature of global markets meant that events in the US fed into events in Britain and Europe, and vice-versa.

A growth in credit availability over the period 1997-2006 allowed the emergence of two linked phenomena: first, persons historically unable to secure mortgages over residential property (‘subprime borrowers’) were able to do so, on increasingly easy terms and at low interest rates. That credit boom fuelled an unprecedented appreciation in residential property values. Between 1997 and 2006, the price of the typical American house increased by 124%. The easy credit combined with what appeared to be an investment that never dropped in value saw people pile into the property market, despite the fact that the market value of the properties they were purchasing was often many multiples of their annual income. As a general rule, when purchasing a house, the value should be no more than three times one’s annual income. By 2006, many mortgages—and almost all subprime mortgages—were taken out over properties valued at five times the annual earnings of those purchasing.

Easy borrowing conditions were generated by large inflows of foreign capital, particularly from the rapidly growing BRIC economies. Broker Peter Schiff called it ‘a giant pool of money’, and pointed out that it demanded relatively safe, income producing investments. Historically, US Treasury bonds were emblematic of this type of investment, but in the early years of the century, US paper did not offer high yields, and there were few similar products. In response to Schiff’s ‘pool of money’, investment banks on Wall Street promoted products such as mortgage-backed securities and tweaked them in innovative ways: whence collateralised debt obligations. Credit rating agencies then assigned the new products high ratings; companies that did not appreciate the risks involved insured them—think, for example, of AIG.

At the same time, American (and British, as we shall see) consumers’ capacity to take on debt had a limit: this was reached sometime in 2007. Thereafter, housing values went into decline. As they did so, major global financial institutions that had borrowed and invested heavily in subprime CDOs reported significant losses.

Britain’s experience formed part of the ‘European contagion’ that spread from toxic US assets, to which British financial institutions had considerable exposure. However, it is important to remember that Britain had also experienced a housing and credit bubble. During the summer of 2007, money markets across Europe became increasingly mistrustful of any institution with considerable investment in mortgage debt. In September, the rate at which British banks would lend to each other—known as the London Interbank Offered Rate (Libor)—rose to its highest level in almost nine years. The first casualty was Northern Rock, Britain’s biggest mortgage lender in 2007 and fifth largest overall, which went cap-in-hand to the Bank of England as lender of last resort on 12 September. Northern Rock had become unable to repay maturing money market loans and experienced a liquidity crisis. Always short on deposits, its subsequent experience of both government deposit guarantees and nationalisation came to symbolise two of the worst sorts of over-leveraging: a low reserve ratio (too little cash physically held in the bank’s own vaults relative to the amount of money loaned to customers) and too much uncertain investment (borrowing on the short-term money market coupled with its own subprime lending, often at 125% LTV).

What happened thereafter has become folklore, as financial institutions and insurers fell like ninepins across the US and UK, banks also stopped lending to business, and governments stepped in to prop up the financial services industry in an attempt to stem what had become a catastrophic credit retraction.

Pointing the finger I: collateralised debt obligations

As should be clear, the crisis was caused by a complex interplay of factors. As to which bear the most responsibility, in my view government policies that encouraged home ownership—providing easier access to loans for subprime borrowers—and overvaluation of bundled subprime mortgages based on the theory that housing prices would continue to rise, are most salient. However, questionable trading practices on behalf of both buyers and sellers, and a lack of adequate capital holdings by banks and insurance companies backing the financial commitments they made are also important. In what follows, I will outline two of the less widely understood causes.

Complex financial instruments that undermined the accurate pricing of risk are central [see Fig. #1 below]. While looking at the diagram, it is important to keep two things in mind: first, it was not understood that mortgage defaults tend to occur in clusters, much like communicable diseases such as influenza or measles; and secondly, there were a number of perverse incentives in the mortgage market—particularly in the US, but also in the UK—encouraging lending to just the sort of people who were likely to default on both their mortgages and other credit obligations.

Banks and other financial institutions bought mortgages from ‘originators’, the companies that first made the loans. The banks then repackaged the loans into ‘pools’, which provided the rights to a stream of mortgage repayments (along with the risks that the loans would go unpaid).  At this stage the repackaging was known as a ‘residential mortgage backed security’, or RMBS. ‘RMBS’ products are steps one and two on the diagram. At each level, the securities were rated as to their riskiness. The buyer of the lowest rated tranche was like the ground-floor owner-occupier in a flood: he was exposed to the most actual losses, but received a higher rate of interest in compensation. The next tranche’s buyer—the owner of the first floor—took the next wave of losses and in exchange took the second highest interest rate, and so on.

The investor in the top floor of the building—with his AAA rated security—received the lowest rate of interest, but this was coupled with the assurance that his investment would not be wiped out.

Had it not been for the credit-driven explosion in subprime lending, this system seemed logical enough: it was meant to create a safe and predictable income stream which could then be traded between banks, pension funds, and other financial institutions. However, when investment banks went about developing new products beyond RMBS (which had existed since the 1960s in the form of mortgage bonds), it is vital to remember that they created CDOs by taking the lowest rated tranches—the floors most likely to flood—and repackaging them. This repackaging is seen in steps four and five. This process was often repeated, magnifying the risk. However, at the same time as the risk was magnified, mortgagors across the developed world—but especially in the US—began to default in large numbers. Risk had become uncertainty as what looked like highly rated assets turned out to be toxic.

The above account illustrates the power of hindsight, and draws on economic analyses that have, for the most part, emerged since 2009. At the time—before 2007—people trading in CDOs did not know their value, did not understand the products, and relied on credit ratings agencies to price the risk. Some were so complex that even if all component parts were identified and valued separately, they still could not be accurately priced.

Pointing the finger II: everyone should own his own home

Widespread home ownership is linked to a number of positive social effects, especially for lower income households: these include enhanced outcomes for children (most notably higher rates of school completion and lower rates of teen pregnancy), anti-poverty effects, and stronger community involvement. In seeking to enhance these effects, government policies in both the US and the UK encouraged home ownership at the same time as the expansion in credit allowed people to borrow more easily. In the US from 2000 to 2003, for example, the Federal Reserve lowered the federal funds rate from 6.5% to 1.0%; the lower interest rate encouraged borrowing.

In the UK, the ‘Right to Buy’ scheme is emblematic of policies that saw many low-income people enter the property market for the first time, while in the US, pressure (and in some cases, legislation) imposed on both quasi-public and private sector lenders caused mortgage lenders to relax underwriting standards and originate riskier mortgages to less creditworthy borrowers [Fig. #2, left]. It is easy to condemn this kind of ‘own your home at any cost’ policy with the benefit of hindsight, and the collapse in underwriting standards at HBOS or Fannie Mae makes much of the subsequent criticism fair. However, in both countries, historic lender prejudice had denied access to a significant route out of poverty for many prudent people on lower incomes. Most notorious was the practice (in the US) of ‘redlining’: marks on a map to delineate the areas where banks would not invest. This was almost always on the basis of race.

Unfortunately, policies to increase levels of home ownership intersected with the housing bubble and its allied explosion in cheap credit. Indeed, some lenders (notably Countrywide Financial) became notorious for ‘reverse redlining’: deliberately targeting low-income customers with a view to signing them up to a subprime mortgage, often predatory in effect. For these customers, keeping their heads above water depended on the value of their property continuing to rise. No investment increases in value indefinitely, and it was often low-income borrowers who finished up ‘under water’ (owing more than the property was worth) when the housing bubble burst.

Preventing a future crisis: industry responses

Most industry responses since the crisis have come about through government fiat: the state has taken control of much of the financial sector and imposed changes from without. However, two internal changes in response to the crisis are salient. The first is a notable return to prudence in lending practices, particularly for residential mortgages, and the second is a new emphasis on training and background checks for senior financial sector employees. More minor—but publicly popular—responses include either sacking or withholding performance bonuses from individuals culpable for reckless strategies during the crisis. There has also been increased oversight from the private sector regulator—the FSA is not a public body.

It is difficult to convey, post-crisis, the nature of loans widely available before it. Many required no documentation from the borrower—so-called ‘self-certification’ loans. Others demanded no deposit and were deliberately lent at overvalue, often with Britain’s burgeoning army of DIY and ‘buy to let’ investors in mind. Others again were interest only; one Countrywide Financial mortgage in the US demanded no repayments as long as the property continued to rise in value—creating negative amortisation (where the amount owed increases over at least part of the loan lifetime). Trader Michael Lewis argues that products meant for sophisticated investors were deliberately marketed to retail customers, who simply did not understand them.

Banks now expect significant deposits—‘ninjas’ (no income, no job, and no assets) no longer qualify and loan variety has shrunk dramatically. Similarly, banks routinely  undertake careful background checks—‘liar loans’ (self-certifications) no longer exist. LTV is often set at 80% and sometimes even lower. Of course, this means it is difficult to get a foot on the first rung of the property ladder, prompting much hand-wringing from social policy analysts keen on the positive effects of home ownership. That said, Australian banks remained fairly conservative in their lending practices in the lead-up to the financial crisis, a practice allied with higher interest rates set by the Reserve Bank of Australia. This had the extraordinary consequence that Australia did not experience the GFC [Fig. #3, left].

Allied to more conservative residential mortgage lending (probably necessary) is the more serious problem of reluctance to lend to even solvent small businesses coupled with the withdrawal of longstanding credit facilities (such as overdrafts). While perfectly understandable in context, this is bad for the British economy, causing serious cash-flow problems.

Preventing a future crisis: government responses

It is fair to say that Britain is still suffering effects from the GFC. While the economy has emerged from recession—as of October 25—there is ongoing retrenchment in the manufacturing sector and over-reliance on service industries. It is important to remember that during the lead-up to the crisis, many sovereign states went on a credit binge of their own and continue to run large deficits: just as banks and households must cut spending, so too must the British government. The recent round of spending cuts goes by the name of ‘austerity’, although it is important to remember that true austerity combines tax rises with spending cuts, and—unlike in Greece or Ireland—Britain has not yet attempted both simultaneously.

‘Austerity’ apart, the most significant government responses are as follows:

  • Prudential regulation: the forthcoming Banking Reform Bill seeks to separate retail from investment banking with a view to protecting important retail functions and avoiding the misuse of customer deposits on speculative investments. Business Secretary Vince Cable’s proposed ‘Small Business Bank’ is meant to address the financial sector’s reluctance to lend. This is positive for business and good for the British economy, but does expose the taxpayer to higher levels of risk (ie, the taxpayer will be taking on loans that High Street banks consider uncertain).

  • Quantitative easing: meant to inject money into the economy without increasing inflation (it does not involve printing more notes), QE increases the capital available for banks to make loans, once again primarily to small and medium enterprises. It has not been an unqualified success: often banks have simply sat on the funds in order to improve their credit ratios.
  • Disengagement from the Eurozone: While the financial crisis was in process, the Eurozone entered a separate crisis of its own, largely to do with the overvaluing of labour and assets in the unproductive ‘PIIG’ economies (Portugal, Ireland, Italy, and Greece) as a consequence of monetary union without sufficient market discipline: in short, the area covered by the single currency was insufficiently homogenous for economic shocks to affect its different regions in similar ways. While Britain stayed out of the euro, much of its trade is still with weak Eurozone economies. This means that Britain is actively seeking more economically robust trading partners. The Eurozone’s financial sector in particular is in such a parlous state that Australia’s (prudently managed) financial sector is now worth more [Fig. #4, right].

Pertinent future issues for our client

Our client is a Scottish bank with Scottish headquarters. This means that—despite the global outlook expected of any large financial entity—there are issues peculiar to its Scottish domicile. The first is the realistic possibility of Scottish independence in the next five years. This may be coupled with a new currency (there is no guarantee that an independent Scotland would keep Sterling, although that possibility is being mooted). Regardless of what currency an independent Scotland adopts, there is also the separate issue of its credit rating, which Fitch broached on 19 October (BBC News):

[T]he rating would depend on the terms of an independence agreement, and how long it takes for the transition to take place. That’s as well as uncertainty on splitting assets and liabilities, and the question over Scotland’s future financial regulation.

The Fitch analyst’s note goes on to point out that an independent Scotland will have no credit history, which means that it may start out without the treasured AAA rating. This would increase its borrowing costs and reduce Scottish banks’ ability to raise finance via corporate bonds.

Should Scotland remain in the UK, the long-term effects of both the Eurozone crisis and new financial sector regulations are still unclear.


  1. Posted October 27, 2012 at 8:59 pm | Permalink

    A glib observation: the diagram of CDO assembly looks a lot like the diagrams I’ve seen of some cryptographic functions.

  2. Joseph Clark
    Posted October 27, 2012 at 9:15 pm | Permalink

    I think this is overall a sensible analysis. I agree with the overall narrative: mispricing of credit together with poor credit monitoring lead to a credit crunch which spread to other markets. In many ways this was a very run of the mill credit crisis. The same thing happens reasonably regularly in every economy with a developed credit market. Sometimes the loans are for houses, sometimes for other speculative assets: gold, tulips, internet companies, infrastructure assets, biotech. The mechanism is the same each time. It’s human to think that good times or bad times will continue to roll.

    A couple of extra comments, and maybe these are a little nitpicky, but economists are like that 🙂

    1. Reserves aren’t usually physical currency in vaults, they’re things like clearing accounts with the central bank, or some kind of low risk liquid debt.
    2. Quantitative easing does cause inflation. That’s usually the point of it, to shift the price level. Whether physical currency is printed or not doesn’t matter. Money is created by the central bank making an entry on their computer system. There’s a few different ways they can do this: they can assets directly (like car companies or mortgages), or increase loans to banks (maybe by decreasing the collateral required), or purchase government debt. The mechanisms are different but if the effect is to increase the amount of money available the intended result is to increase asset prices.
    3. The credit rating of the government doesn’t necessarily affect banks ability to raise capital. If the bank has good collateral and makes good loans people will be willing to lend to it. If the market is worried that the government will seize bank assets or screw up the economy it could affect bank financing costs but that’s a pretty extreme case.

    A final point of disagreement. It is said a lot that people didn’t understand the credit instruments and this caused them to price incorrectly. I started working in a fixed income team in 2007 and there was a very acute understanding of how things could go bad if the underlying assets went bad. There might have been some people who invested in CDOs without knowing how they worked but not many. They’re really not that complicated. People who held them were basically making a bet that prices wouldn’t go down too much. They were wrong, maybe a little silly, but it’s easy to say that after the fact. Everyone had an opportunity to put a bet against it. If you correctly predicted the extent of the crisis you could have easily turned 10k in to 10m. A lot of people talk about how obviously silly it all was but few of them backed their view with their own money. That kind of talk is cheap.

    Anyway thanks for the good read.

  3. kvd
    Posted October 28, 2012 at 2:59 am | Permalink

    This is a good analysis of what happened from the lenders’ point of view – where they exceeded what should have been prudential limits, etc. but fwiw I’d like to turn it on its head, and suggest that you could also look at the ’causes’ from the borrowers’ perspective, and reach a conclusion more closely linked with industry and employment policies.

    1. When you say ‘the value of the property should be no more than 3 times annual earnings’ I would rephrase that as the ‘loan undertaken should be no more…’. This gets to capacity to service the loan, irrespective of the value of the property. And further, if capacity to repay is maintained, the movement in value of the property is then irrelevant to the lender except in the case of default.
    2. Defaults obviously did occur, but not because of a ‘bubble burst’ in the value, more because borrowers were unable to maintain repayment obligations, for two reasons: a) unreliability of continuing employment; b) imprudent borrowing limits.
    3. If 2 above is fair, then government failure and government response should be directed to those two factors, and you have correctly noted increased prudential regulation as a part of this, but I’d suggest that equally important is the government’s response to maintaining or increasing, the employment opportunities for its population.

    Take the above to extreme. CDOs are no more than future repayment streams. If government policy and economic imperatives are such that you place pressure on reliable employment leading to significant loss of national employment, then those income streams will fall, and defaults become inevitable.

    Wouldn’t matter if you ‘prudently’ allow a borrower to access only 10% of his income against a maximum of 10% of the property value. If he loses employment, then he cannot repay, and the whole house of cards falls down. The ‘pass the parcel’ lending practices in the US UK and Europe only really came to a halt when their populations lost the capacity to service over-egged borrowings through falling employment.

    So if I were Scotland I would look long and hard at industry policy which provided and promoted employment opportunities for its population.

    And like Joseph above, thanks for a very good read.

  4. TerjeP
    Posted October 28, 2012 at 4:16 am | Permalink

    A great analysis and unlike Joseph I wasn’t familiar with how CDOs were assembled, at least not in such detail, so this bit was very helpful.

    A few comments:-

    Quantitative Easing (QE) may not literally entail “printing notes” but it is qualitatively so similar that the term “printing notes” is a fair and reasonable simplification. Although in this context I’d personally say “currency” instead of “notes”. And it would be inflationary if the recipients were not “hoarding the cash”. It will need to be reined in when velocity returns or else it could be very inflationary.

    Missing from the discussion is the role of non recourse loans. Although I’m not sure of the real significance of these.

    The distinction between risk and uncertainty is quite profound. It is interesting that when the financial sector was faced with uncertainty it sensibly limited its exposure but many governments assumed they knew better. I’m thinking of the Irish government in particular that paid heavily for guaranteeing the banks. Although the Irish example is most notable simply because it lost the bet. The Australian government made much the same bet but I doubt it was any better informed and suspect it was merely luck rather than sound judgement.

    Has risk analysis in the industry changed in practice to now account for cluster style events? I would think the insurers should be all over this issue.

  5. Posted October 28, 2012 at 5:12 am | Permalink

    Hi SL. A companion post – http://belshaw.blogspot.com.au/2012/10/sunday-essay-australia-economic-luck.html

  6. Posted October 28, 2012 at 6:22 am | Permalink

    Great post, though it is a pity I had not quite finished a relevant post! (Hopefully, it will be up in a day or so).

    On the housing price surge and crash — the role of land rationing is significant. (This is particularly obvious if you look at where housing price booms did and did not take place in the US.) If supply cannot respond smoothly to increased demand then prices rise. If this happens persistently, then the expectation of capital gains are set up and will continue as long as credit is available and being used to fuel it. After all, if supply responds smoothly to demand, the price rise pattern does not set in in the first place. (Remembering that, as Matt Yglesias puts it, houses are large decaying physical objects; it is the land they are on which shoots around in value.)

    The problem with QE1 and QE2 is precisely that they were explicitly terminating, so did not change expectations (hence banks sitting on cash: the Fed paying interest on reserves does not help either). One notices that the Fed has “got the message” and announced that QE3 will continue until unemployment has fallen. (Like other market monetarists, I dislike a central bank aiming for a “real” target; but the clear aim is better expectations management.)

    I would also add in clear expectations management by the RBA — relatively quickly, people regained confidence that spending (and so incomes) would not crash. That Oz policy was coordinated, quickly responsive and clearly had lots of room to manouevre was also helpful on the financial side.

  7. Posted October 28, 2012 at 7:08 am | Permalink

    Lorenzo: Jim makes the same point in his excellent post (linked above @5) that you do in your last par.

    Inevitably there will be abbreviations and amendments, but I am glad people I respect are taking something from this, and I am particularly proud of my explanation of RMBSs/CDOs (so thanks Terje!).

  8. Posted October 28, 2012 at 7:43 am | Permalink

    Thank you, Helen. It takes time and effort to write with the clarity you did. It enabled me to zero in on the key point that I disagreed with and then, hopefully, explain my disagreement with something approaching your clarity.

    One of the things that I really love about high level blogging, and you are all high level bloggers, is the conversation that results. This includes the regular commenters on this blog. I don’t always comment myself, but I do learn. I reckon that’s pretty good!

  9. kvd
    Posted October 28, 2012 at 7:50 am | Permalink

    Yes, I was most remiss in not congratulating you on the CDO explanation. Probably one of the clearest of the many I have read. So, well done, again.

  10. Posted October 28, 2012 at 7:56 am | Permalink

    In terms of clarity, it’s worth pointing out that when I wrote (w.r.t. quantitative easing) that it was ‘meant to inject money into the economy without increasing inflation’, I was hoping that the ‘meant’ would flag to the world that I didn’t think this was always or necessarily going to be true. Clearly the single word wasn’t enough, so there are clarity lessons everywhere.

    I await Lorenzo’s follow-up post keenly.

  11. Posted October 28, 2012 at 9:13 am | Permalink

    Actually, having slept on it, the nature of CDOs goes into one of my pet peeves, which is the destruction of useful information. I wrote about that a bit in the context of Yet Another Book Review.

    Basically, the chopping and rearranging, then re-chopping and re-arranging, through multiple iterated rounds, effectively destroys useful information about the actual origin of funds.

    We live in an age of electronic brains that are very good at tracking staggering amounts of minutiae. There’s no technical reason that the providence of every tranche couldn’t be traced all the way back to mortgagees and vice versa.

  12. Posted October 28, 2012 at 9:15 am | Permalink

    And behold, chopping and swapping really is one of the fundamental pillars of encryption.

  13. hhoran
    Posted October 28, 2012 at 9:28 am | Permalink

    A fascinating, thought provoking exercise. I am very appreciative that you’ve shared this given the surprising lack of basic narratives on the crash. The worthwhile task you defined was to provide a high level definition of the problem and a central thesis about causation, recognizing the challenge of untangling and prioritizing the many overlapping “causes” and “problems” that any historical event of this magnitude involves.

    I have no issues with any of the factual points mentioned, but I would seriously dispute both your definition and central thesis. Any definition of the crisis needs to focus on the creation of massive systematic financial risk (in conjunction with the weakening of historical defenses against systematic risk becoming widely contagious). Much of your paper discusses aspects of this, but “systemic risk” is never explicitly mentioned. The liquidity contraction you cite was a symptom of this risk, not the central event of the crisis or its primary cause.

    Similarly the “housing bubble” must also be seen as a symptom of more important problems, and not an exogenous variable that came out of nowhere. Housing markets have always been subject to cycles of booms and busts, but could not have contributed to massive systemic financial risk in the absence of other more critical structural changes. The “housing bubble” was a nit compared to the “financial bubble” where the FIRE sector grew from 4% to 16% of the US economy without any plausible relationship between that growth and new products/services that provided real value to anyone in the real world.

    Systemic risk results when many of the historical constraints that would limit the FIRE sector to activities that created legitimate economic value broke down as the sector pursued mind-boggling levels of short-term gains and the political power needed to fuel those gains. Basic break downs included capital reserve requirements, strict links between risks and returns, protections against fraudulent financial products, the corruption or elimination of adversarial safeguards (regulators, rating agencies, barriers against contagion from “casino-like” speculation) and many others that you and your readers will be familiar with.

    The destructive real-estate financing you mentioned would never have become this destructive without 30 years of changes that unleashed an entire sector in the fanatical pursuit of outsized returns that were largely wealth transfers from other sectors, instead of net contributions to overall economic growth. If the problem had started when Countrywide and similar rogue firms abandoned lending standards, you would have seen a bunch of bankruptcies (similar to the dot-com collapse, or the 1980s S&L crisis) and you would never have seen systemic risk spreading globally. Congress did not relax housing lending standards because civil rights activists finally convinced them that redlining was bad. The relaxed them because financial industry lobbyists flooded Congress with demands to eliminate protections against predatory lending (and BTW Fannie and Freddie only got increased freedom years after everyone else had gotten it).

    Housing (for the historical reasons you cite) just happens to be the place where the financial dam first burst, but the financial sector was pursuing similarly destructive behavior lots of other places. Your paper implies (but doesn’t clearly state) that the crisis was global, and covered lots of places where there was no housing bubble. There were destructive financial practices in lots of areas other than US/UK/Spanish housing. Lending practices in the Eurozone. Predatory private equity practices, where the productive capability of the economy gets crippled in order to strip out assets and cash. M&A practices driving consolidation in dozens of industries that don’t need it, destroying future innovation and resiliency in order to strip out more cash. The dam was going to burst at some point. The “problem” isn’t the bit of concrete that fails first, but how uncontainable pressures were built up in the first place, and why all the protections against those pressures had been deliberately dismantled.

    Having argued that the “housing bubble” was just a subset/symptom of the much larger “financial bubble” I would acknowledge that it would probably also be fair to point out that the “financial bubble” was a subset/symptom of a larger “rentier bubble” whereby certain political/class interests are expropriating short-term value from other sectors/classes. Stuff that’s been going on for centuries in various forms, but has reached previously unimaginable levels of magnitude and risk, given globalization and other movements. There are lots of destructive rentier movements outside the FIRE sector (i.e. military suppliers, intellectual property) that are certainly amplifying the effects of the financial crash and making recovery more difficult. The question is whether is whether it is more useful (if one is attempting a “simple” explanation of the crash) to focus strictly on finance and systemic risk as a central bank is supposed to, or to focus on the political movements that drove much more than the financial changes.

    P.S. Again thanks for the original post, and yes, I understand that your client might not be terribly receptive to issues broader than the “housing bubble”.

  14. kvd
    Posted October 28, 2012 at 9:47 am | Permalink

    [email protected] one of us is misunderstanding – usually it’s me, but here goes…

    SL’s point (or one anyway) was the failure of adequate risk assessment. From her very good diagram you should be able to tell two basic things:

    1) That tracing a CDO back to underlying security was/is most certainly possible, but
    2) The further from that reality the onsold securities became, the more attention/significance was given to the ratings agencies improper assessment of what were essentially BBB & BB- securities as AAA-rated in that revised reincarnation.

    In Oz our banks have recently brought in a modified form of this sort of securitisation, but significantly are applying their process over a ‘moving basket’ of underlying mortgages, as opposed to specific groups of mortgages, thus they are retaining/carrying the risk of default, and assessment is therefore made much more transparent.

    So, while you are correct that it was ‘hard’ to ascertain status, I think it remains true that it was the improper risk assessment which attached AAA ratings to what were/are grossly non-performing loans.

    Or I could be completely wrong, and would therefore welcome a different explanation 😉

  15. TerjeP
    Posted October 28, 2012 at 9:47 am | Permalink

    SL – did you make the graphic explaining CDOs? It really nailed it. And now that I’ve got a better grip on how they are constructed I’m a lot less cynical about them. Essentially they are just an attempt at risk reduction through diversification. Quite reasonable in theory except that the assumed risk model was probably wrong and the way the risk of a subsequent portfolio is calculated was also possibly wrong. The scope for synchronised failure not being sufficiently factored into the equation.

    p.s. Josephs point about people being wise after the fact is also important. Where were the critics prior to the crash and did they put their money where there mouth was?

  16. TerjeP
    Posted October 28, 2012 at 9:51 am | Permalink

    p.s. my ability to use the word “their” and “there” correctly and incorrectly in the same sentence, as per the last sentence above, truly does confound me. My brain was not wired for written language. Doh!

  17. kvd
    Posted October 28, 2012 at 9:59 am | Permalink

    [email protected] see here, in the section headed “Financial innovation and complexity”. I agree with the rest of your comment.

    Down here on the ground, we have my local council’s finance committee happily invested in these dodgy securities, all on the basis of a AAA rating from one of the ratings agencies. That is as far as they looked, as I understand it. Court actions are in progress to try to pin blame somewhere, anywhere else at the moment.

  18. Posted October 28, 2012 at 11:48 am | Permalink

    Jacques, yes it really does look like an encryption process. It’s almost like an attempt at fraud by exploiting the informational or computational limits of your victims, rather than through direct misinformation.

    I think one thing that regulators need to learn from the CDO experience is that a mark to market approach, by itself, is a grossly inadequate to minimise systemetic uncertainty. You’ve got to go deeper than valuing reserve assets based on what some idiot will pay for them today, and analyse what they’ll be worth in a time of crisis.

  19. Steve Roberts
    Posted October 28, 2012 at 11:57 am | Permalink

    The aim of a cryptographic function is to mash up the input data beyond all recognition and recall, unless the reader possesses the secret plumbing that can un-mash it. (Hash functions are very similar, and have no way back for anyone). So the observation is that this processing of money, whether by design or by coincidence, is obscuring where the money came from originally. Hell of a paper, by the way Helen, well done.

  20. Posted October 28, 2012 at 2:19 pm | Permalink


    Just to nitpick, useful crypto is built on Kerckhoffs’s Principle: an algorithm should be secure even if the attacker knows exactly how it works.

  21. Posted October 28, 2012 at 3:54 pm | Permalink

    [email protected] I was actually agreeing with JB’s comments.

  22. Holden Caulfield
    Posted October 28, 2012 at 4:14 pm | Permalink

    This is a nifty summary, especially the diagram. But the most important point made is the crucial distinction between risk and uncertainty, which, as you note, is that the former is quantifiable. The reason this is so important is if you can quantify the risk of an investment, you know how much insurance you need to buy to make the investment risk-free.

    Investment bankers had been salivating over mortgages for years due to the sheer size of the market. Securitization was just the logical step from the insight that the more diversified an investment portfolio became, the lower its risk; or don’t put all your eggs in one basket. They had already securitized corporate bonds, car loans, and even education loans.

    However, given the heterogeneous nature of residential mortgages – duration, interest rate, location, borrower’s credit rating – they hadn’t worked out how to calculate the risk of a bundle of residential mortgages, compared to just one. To do this, they had to calculate the correlation risk of two different mortgages going bust, and then extend that to the correlation of 100, 1,000, and so on.

    One guy published an article claiming to prove an algorithm (copula as it is known in statistics) that quantified the dependence among all these different mortgages, and thus the probability they would all go bust at the same time (or a certain % of them). This copula thus provided a single risk calculation for a whole bundle of mortgages, and thus tranches of that bundle could be priced, and, more importantly, hedged. Or so they thought.

    A major lesson learnt here is that investment banks need to start hiring more former real estate rental managers and junior mortgage assessors, and fewer PhDs in Finance and Statistics, as the latter were clearly clueless about what risks actually exist in the real world, particularly in its most mundane processes, like buying a house, and keeping up with mortgage payments.

  23. Posted October 28, 2012 at 4:54 pm | Permalink


    Wired had an article about the famous Gaussian copula formula.

  24. Holden Caulfield
    Posted October 28, 2012 at 5:01 pm | Permalink


    I actually worked with it briefly. Like tonnes of other people, I knew it was crap. The thing is if you are just a salaried (plus bonus) employee, you simply don’t care. All that matters is if everyone else believes in it, and puts their money where their mouth is, it is reality. To those who knew what was going on, the way to make money was to think of the industry as one huge game of musical chairs.

  25. Posted October 28, 2012 at 7:51 pm | Permalink

    Terje: no, I didn’t make the diagram, it’s from the IMF (which if you click through and blow it up a wee bit, you can see down in the bottom left hand corner). However, I found it as an orphan graphic without any accompanying commentary (which is a pity, because even as a stand-alone it’s very good). Initially I was going to write a septic tank analogy (mainly because the diagram looks like plumbing/drainage), but not a lot of people in the UK have septics/bush loos – that’s more an Australian/US thing. I would have finished up having to explain two things at once. Plus the sewage imagery is fairly gross.

    Before I found the diagram I was going to use Tim Harford’s ‘how great is the chance of getting a rotten egg in each carton?’ + ‘what happens when you repackage the eggs in different cartons? Does the risk of landing a rotten egg go up or down?’ from his column in the FT. However, that would have involved me explaining the maths that Holden mentions (yes, I have a pair of high distinctions in a statistics course and a finance course…), and I wanted to avoid that, because maths just makes a lot of people’s eyes glaze over. So I went with a plumbing/flooding analogy, which in combination with the IMF diagram seems to have worked.

    And as hhoran points out @13, there is a lot more I could have included, and I do regret not being able to get the S&L crisis in there (I really would have liked to mention the moral hazard attached to bailing out expert rent-seekers). However, Ron Paul and Gary Johnson have both been pretty good at getting the message out about moral hazard (it’s been widely reported over here, too), so I thought if I had to drop something, it could be that.

    I also read a lot of criticism of policies intended to increase home ownership rates, without any understanding as to why governments/local authorities/private bodies thought that was a good idea at the time, so I thought that particular issue needed addressing in context.

  26. Andrew Reynolds
    Posted October 28, 2012 at 9:12 pm | Permalink

    A well reasoned analysis. Just one correction. As many do, at one point you confuse capital and liquidity. You say “…QE increases the capital available for banks to make loans…”.

    QE can take several forms. Mostly (as in the US) it involves printing (or creating computer records of) money and then lending that to banks at low rates of interest in the hope that they will then lend it out. This process creates no capital at all, just liquidity. It can only create capital slowly (if the banks then make profits) or, in extreme circumstances by the government directly buying shares or other capital instruments of the banks.

  27. Posted October 29, 2012 at 6:13 am | Permalink

    [email protected] I should say, i was agreeing with Jim Belshaw’s comments except his last paragraph in his post.

  28. Mel
    Posted October 30, 2012 at 1:58 pm | Permalink

    Some comments on the OP:

    “As a general rule, when purchasing a house, the value should be no more than three times one’s annual income.”

    Is that house value or the loan amount? I also find the three to one figure hard to swallow. Most Australians have borrowed at a way higher ratio without a concerning level of default. I would go further and say that a three to rule would be an economic and social disaster. Most of us would end up living in trailer parks or small mass produced concrete high rises under such a vision. Or in rental. For ever.

    “It is important to remember that during the lead-up to the crisis, many sovereign states went on a credit binge of their own and continue to run large deficits: just as banks and households must cut spending, so too must the British government.”

    Why conflate these points in the one sentence? Are you saying banks, households and government ought be cutting spending *at the same time*? This is a hugely controversial call. If so I think you should state what economic theory you are working with and at least write a para explaining why the still mainstream Keynesian arguments are wrong.

    I have many other issues with what you’ve wrote and what you’ve excluded from consideration. Ultimately, however, this issue is complex and there is no definitive “right” answer to the cause of the GFC. I’ve read many supposedly definitive accounts and I won’t pretend that I have the smarts or the background knowledge necessary to synthesise these into a complete picture. Your account is as good as many others, I suppose.

    I also don’t understand why a legal expert such as yourself would be asked to write such a brief. Surely this issue is technical enough to require someone with the appropriate specialist training, namely an economist with suitable relevant expertise. Or maybe these types of briefs are too inconsequential for it to matter.

  29. Posted November 3, 2012 at 2:40 pm | Permalink

    I read this post a while ago and again; I generally refrain from commenting on blogs at the moment so I didn’t. But there’s an itch that needs to be scratched unfortunately so here it is…

    This post and some of the following commentary, learned and accurate as they are, betrays the basic flaws in economic thought. For example:

    The mechanism is the same each time. It’s human to think that good times or bad times will continue to roll.

    It may sound pedantic but the mechanism is not the same, there is no mechanism. There is a pattern and, altho’ the causes are multifarious, the pattern and its underlying psychology recurs. Joseph Clark is right in that it has something to do with human nature and its recurring faith that good times will roll on. But he pattern is not a mechanism, rather it uses the complexities of finance to create a mechanism which promises great wealth in a short time for anyone who wants it.

    I fail too see how the mere existence of the sub-prime market is a major factor here except insofar as it allows people of limited means to believe however briefly that they too will be greatly wealthy. What is really striking is the mechanism illustrated in the first diagram. I am instantly reminded of the Blue Ridge Corporation, the Shenandoah Corporation, Harrison Williams and Goldman Sachs.

    I understand the differences but they do not explain away or qualify adequately the similarities. A pyramid is created, at the top there appears a solid investment but it a marble statue whose feet are clay. It’s a bet that those who will probably default on their mortgages and lose their property in the event of an economic downturn, mostly won’t. And it’s exacerbated by the ridiculously high value placed on this property.

    The situation is simple and any reasonably sensible adolescent could see that. for example, this particular flower shouldn’t be worth more than a four bedroom house, or that this company is not producing anything people want to buy despite the highfalutin digital technology schpiels of its executive leaders.

    Yet, due to particular interests and ideological dispositions, entire industries are created around arguments about the causes of such crashes. People went nuts. Why? because they don’t know their history. And in their ignorance and their rush to ride high on the boom and get rich they make the same mistakes over and over again. And when they do the government steps in because that is what is demanded.

    Simple. But unpalatable.

One Trackback

  1. By Skepticlawyer » The year in review on December 30, 2012 at 1:01 am

    […] heads lest we get overheard saying something untoward. I also turn (not having done it for a while) to the economics of the GFC, while Lorenzo provides a thoughtful […]

Post a Comment

Your email is never published nor shared. Required fields are marked *