I wrote this in late September 2011, to explain to my circle of friends why I thought we were in the state we were in. It’s by way of background to my latest post on secular stagnation, so I’ve disabled comments on this one.
The Global Bezzle: Whence it came, where it went and what it means
By way of a reply to both John Quiggin and Brad DeLong, below I set out my view (which is largely based on, by which I mean largely stolen from, Dean Baker’s in “The End Of Loser Liberalism“, including one big change of position; obviously Dean isn’t responsible for my presentation) of the roots of the current crisis, with particular evidence on how the financial sector was really quite irrelevant, other than in performing its normal role of intermediating much bigger imbalances and policy errors …
The global bezzle
The term, from JK Galbraith’s “The Great Crash”, originated in a quote about embezzlement:
“In many ways the effect of the crash on embezzlement was more significant than on suicide. To the economist embezzlement is the most interesting of crimes. Alone among the various forms of larceny it has a time parameter. Weeks, months, or years may elapse between the commission of the crime and its discovery. (This is a period, incidentally, when the embezzler has his gain and the man who has been embezzled, oddly enough, feels no loss. there is a net increase in psychic wealth.) At any given time there exists an inventory of undiscovered embezzlement in – or more precisely not in – the country’s business and banks. This inventory – it should be called the bezzle. It also varies in size with the business cycle.”
But I think it’s more normally used these days in a generalised sense – simply referring to that proportion of national (and indeed global) wealth which is made up of illusory, unsustainable and (obviously) unrealised capital gains. And the global bezzle matters as an economic phenomenon because people borrow against it, creating burdens of nominal debt which overhang the economy and which deepen recessions. The bezzle, at its peak value in 2006, was historically large, which is basically why we’re in the state we’re in now.
Whence it didn’t come
Given the amount of money used in financial sector bailouts, it is overpowering tempting to assume that the bezzle must have been located somewhere in the financial sector, possibly in subprime CDOs or derivatives or something too complicated to understand. There is, as I’ve said a couple of times before, a powerful confluence of interests in portraying the crisis as being something incomprehensibly technical in nature. Actually, it was mainly house prices. It had to be.
It couldn’t have been the subprime market, it’s not big enough. At the peak in 2006, subprime mortgages in the USA totalled about US$1.5trn in total outstandings. This is slightly less than 15% of the total mortgage market, and somewhat less than 10% of total housing values. The S&P500 alone is capitalised at a bit more than US$10trn; the average monthly fluctuation in equity market values as a whole is somewhat more than the total value of the subprime mortgage market. Added to which, you will note that the “financial crisis”, with the notable exception of AIG (and of course, AIG is a “notable exception” in the non-pejorative, rule-proving sense, because it was brought down by non-insurance trading and securities lending operations) was more or less purely a banking crisis. Life assurance, pensions and mutual fund companies, to an overwhelming extent, were not exposed to subprime bonds; the main owners were hedge funds and the banking system itself. Given this, it seems unlikely that subprime could possibly have caused a large enough wealth effect to slow the economy down so badly. The US$6trn of housing wealth lost by Americans seems a more likely candidate.
It couldn’t have been anything specifically American, it happened in too many different places. Geographically, within the USA, the severity of the recession matches up to the house price rise and fall, not to the incidence of subprime lending. Spain had a very severe downturn, despite having a very different mortgage market; Ireland’s was different in another way, while Greece hardly saw a lending boom at all. Australia has seen lending practices aggressive enough to strike fear into the hearts of Americans for years (a credit card linked to the mortgage balance, cobber?), while Canada saw its mortgage securitisation market freeze up rather more totally than the American one did.
The financial sector’s involvement was demand-pull, not supply-push. We can contrast the dot com bubble with the telecoms one that happened at almost the same time. In both cases, the financial sector was deeply involved (it is hardly possible to have an investment bubble without them, after all). In both cases, later investigations revealed considerable unethical practice and many cases of out-and-out fraud. In the case of the dot coms, however, it was largely responsive – the public had a more or less insatiable demand for Internet company securities, and the banks were left scrambling to do what they could to satisfy the demand. The telecoms bubble was very different; there was no rabid enthusiasm on the part of the public to fund expensive and predictably unprofitable fibre-optic networks, and the banks had to use every trick in the book to get them done. I would call one a paradigm case of demand-pull and the other a paradigm case of supply-push; in the case of Pets.com, there was nothing manipulated about the absurd valuation, the company was exactly as unprofitable as it reported itself to be, but in the case of Enron, WorldCom and Global Crossing, the valuation was reverse-engineered and manipulated to the levels it needed to reach in order to justify the transactions that Skilling, Ebbers et al wanted to carry out.
The real estate boom, even the subprime element of it, looks a lot more like dot com than telecom. We can note from the dot com boom that there is nothing inconsistent between classifying a speculative bubble as demand-driven, and there being substantial amounts of fraud, hype and aggressive marketing on the part of the financial sector. There were boiler rooms and chop-houses pushing stocks in the dot com years, plus hourly television ads for online brokerages, along with the (absurdly undignified) birth of the “celebrity stock analyst”. But it really is rewriting history to say that dot coms were foisted on an unwilling American investing public by a ruthless financial sector.
The timelines are all wrong. The mortgage bubble was like dot coms. One of the fascinating facts is that, according to three separate teams of researchers who have looked at the LoanPerformance data (a file containing details of roughly half the subprime mortgages ever written), the “massive decline in underwriting standards” for subprime loans never happened. Depending on how you measure it, you might conclude that there was a more or less imperceptible weakening, a more or less imperceptible strengthening of standards, or nothing at all, but nobody has yet found anything to explain the difference in observed default rates between the 2002 and 2006 vintages of subprime loans, other than the fact of the real estate collapse itself.
This is obviously somewhat surprising to anyone (like me) who read the SEC’s complaint against Angelo Mozilo, but it is notable that the complaint (which was settled out of court – Mozilo admitted charges of insider dealing but not the ones of misrepresentation which are relevant here) is very keen on describing frightening products like “exception loans” or “80/20s”, but much less so on attributing actual numbers to the amount of business written on these terms. The two Fed teams and NERA show a picture of the data which is not very consistent with a sudden decision to abandon lending standards in order to feed a securitisation machine, but very consistent indeed with consistent lending standards being hit by a massive exogenous wall of demand, created due to a housing bubble. Talking about timelines, it is worth noting that more or less anything that happened after 2006 (the peak of the housing market, and the beginning of the destruction of housing wealth) was more about passing the hot potato about than anything else. The famous “Abacus” transaction which starred in the Levin Commission’s inquiry, for example, was certainly a very bad deal, but it couldn’t have been a cause of the crisis simply based on the date when it happened – it represented a last-ditch attempt to redistribute losses caused by a crisis that had already happened.
Whence it came
So where did the bezzle come from? In my view, it’s a mistake to see the mortgage and subprime bubble as a separate entity from the dot com/telecom bubble, and to see either as a distinct entity from the general increase in both personal sector indebtedness in the English-speaking economies, and government indebtedness in Euroland. In the past, I’ve emphasised one aspect of this – the real estate bubble as a policy response to the dot com crash – but it’s actually somewhat larger than that. Although the chart of overall personal sector debt has something of a visible inflection point around 2000, you can see that neither the debt service to income nor debt balance to total income charts for the US household sector really show either event. Although there were undeniable aspects of a speculative bubble to both the dot com and real estate bezzles, the actual long term dynamic here – the build-up of overhanging debt on the US private sector and Euroland public sector – doesn’t really look like a bubble. It looks like the cumulation of a long-term and persistent imbalance, ratcheting up the gross debt, which was sustained for longer than it could otherwise have been because of the two speculative bubbles (which inflated the value of the assets on the other side of the balance sheet), and which has now become critical as the asset side has bezzled away, making the net debt position approach the gross debt.
(Wynne Godley was writing about this trend as far back as 2000, before the dot com boom had even peaked, by the way, this point thanks to Oliver “@maxrothbarth” Rivers on twitter).
And it should be clear what we’re talking about here to anyone who has paid even a bit of attention for the last twenty years; the Great Trans-pacific Imbalance. The “savings glut”, the “China effect”, what have you. Basically, the consequence of a) Chinese (also Asian, also to a limited extent other emerging markets) decision to run a large trade surplus as part of a strategy of industrialisation (or because they wanted to be sure never to end up in the position of 1998 Indonesia, or something else), combined with b) the decision on the part of Europe and the USA to accomodate this policy through substantial real appreciation of their own currencies.
If you are importing capital, then the foreign sector surplus has to show up as a deficit in some other sector, mathematically. In the USA it showed up as a deficit in the personal sector, with the increase in indebtedness financing an asset price inflation. In Euroland, it showed up mainly as a big increase in government sector deficit (which in turn piled up in the peripheral European sovereigns, which also ran structural bilateral deficits with Germany), which financed ten years of the Greek policy of “low tax collection and a generous state”, although Spain also had a real estate bubble.
There is a temptation to say that this, in some way, reflect power structures – that in the USA, the political importance and power of the financial sector made it highly likely that the eventual destination of the trans-Pacific capital flows would end up in an asset bubble, while the different power structures of Greece made it more likely that they would be channeled to the locally more politically powerful clients. That’s tempting, and it might even have been right. But it seems more likely to me that things just followed the path of least local resistance, and that in each country, the natural and easiest borrowing sector ended up picking up the deficit. On this analysis, Germany, Australia and Canada came out OK (so far, and Australia has actually had a big run-up in personal debt) because for one reason or another, they had a large enough surplus on their own export trade to offset the capital account imbalance.
But whatever else is the case, given what happened to the balance of trade, it *couldn’t be the case that some sector of the US and European economies didn’t end up building up a large debt balance. The only way this could have happened would have been if the banking sector had taken over the reins of monetary policy, and neutralised the capital flows by lending overseas to surplus countries. Since there is no truth to the belief of “doctrine of immaculate transfer” (the belief that international capital flows can be equilibrated without price movements), this is equivalent to assuming that the banking sector would not only have been able to carry out countercyclical monetary policy by tightening its lending criteria in a boom (an act of damping feedback that is pretty steep to demand in and of itself), but also have been able to carry out a currency policy diametrically opposed to the one that the respective central banks wanted to run. I hope everyone can see why this wasn’t going to fly, and furthermore (I will try to come back to this), does this really sound like the sort of thing you want the banking sector to be doing as a matter of course? Isn’t that what central banks are for? Basically, the job of the banking sector is to intermediate. If it is intermediating a wholly unbalanced state of affairs, it is going to get wholly unbalanced itself. If the Western world is going to assume that it can consume cheap Chinese manufactured goods indefinitely (via a massive structural undervaluation of the yuan), then that is a “real bezzle”, which is going to have a counterpart financial bezzle somewhere.
On this view of the world, the global “financial” crisis had a dress rehearsal in the UK housing market bubble and crash of the 1980s/90s. What we saw there was basically a dress rehearsal for the next twenty years, as the UK decided to hold an overvalued currency (as a form of anti-inflation policy), while accomodating an aggressive trading partner (Germany)’s desire to run a large bilateral trade surplus (its response to weak domestic demand). This was when Nigel Lawson’ doctrine of the “benign and self-correcting” nature of current account deficits, as long as they resulted from private sector imbalances and did not have a government sector counterpart. It didn’t work, for exactly the same reason, which is that the private sector can’t do monetary policy in a deregulated economy (in an economy with capital controls, it of course has no option).
Where it went
So what happened to the money then? Well, in my view it went on three very big defaults. But before I name the defaulters, let’s think about how the hell it was that this was all allowed to happen. Why weren’t people scared? Why didn’t the financial system seize up long before the bezzle got so big? Specifically, why did it go on intermediating for so long?
Well, basically, it was tacitly encouraged to do so. There were three big put options. Call them the “Greenspan put” (large asset value falls will always be met by relaxation of monetary policy), the “Market Liquidity Put” (wholesale funding will always be available) and the “EMU Put” (all European sovereigns are implicitly jointly and severally guaranteed).
They all three had a purpose. The Greenspan Put was meant to support the increase in gross debt of Americans, by ensuring that it would be roughly matched by an increase in the nominal (unrealised) value of asset. The market liquidity put is just an expression of the general Bagehot principle of last resort lending, but it was meant to allow the domestic banking sector to finance the increase in personal sector debt, given that personal sector deposits did not grow at the same rate. And the EMU put was needed to make EMU happen.
And as the bezzle grew, the implicit liability grew and grew, and people got more and more concerned about the fact that because these things were never set out in so many words as a proper agreement, nobody had really checked that they were absolutely happy with writing an open-ended cheque. But meanwhile, the world needed to turn; the container ships from Asia kept on arriving, and consequently the bills kept piling up and the deficits needed to be intermediated. And so the financial sector leant harder and harder on the three big puts. And in my opinion it was pretty fogiveable of them to do so. The market liquidity put, most of all. It absolutely was a reasonable assumption for, say, Northern Rock to make that if they were faced with a deposit run which had little to do with the quality of their assets, but which was forced on them by the shutdown of the US market, and triggered by an irresponsible leak from the state broadcaster, that the central bank would perform its function as a central bank. It turned out to be wrong, but given that this is what central banks have been doing since time immemorial, I think it’s real hindsight to say (as the inquiry report actually did) that it was completely unreasonable to rely on it continuing into the future. The EMU put was really just as respectable, because it was heavily trailed by the political classes of all the Euro countries, and to a significant extent still is. The Greenspan put? Much less respectable. But much less important, too; intermediation didn’t really depend on it.
It’s also worth pointing out that there is a huge winners’ curse here. In the USA, technical aspects of financial regulation meant that chartered banks could not make use of the market liquidity put. Which is why most US subprime lenders were nonbank specialists, who funded themselves by selling paper to the (nonbank) money market mutual funds. This is what the phrase “shadow banking” really signifies. It makes no sense to talk about the shadow banking system as if it was independent of what banks are allowed to do; the extent of the “shadow” system is defined entirely by what functions (economically) the banks have to provide, versus what they are allowed to do. And to hammer home a point, if there is a massive economy-wide imbalance that needs to be intermediated, it will be. And the question of who ends up intermediating it is decided by an auction in which the most aggressive player wins, not the most sensible.
This is where the bezzle went. Assets are not guaranteed by the US government, even though people kind of thought that they would be. The financial system cannot rely on unlimited central bank support, even though it was kind of understood that they could. And the peripheral European states are not guaranteed by the system as a whole, even though for a very long time lots of people were allowed to operate on the assumption that they would be. While we thought that these things might be the case, we were all a lot richer; that’s the definition of a bezzle.
What does it mean?
SO … here we are. We’ve got a lot of manufactured goods hanging round the shop, which in one way or another we are not going to end up paying for – they have a counterpart in claims on our economies by Asian economies, which are denominated in currencies that are not going to be worth as much when translated back as they might previously have been. If one was feeling cute, then one might say that we are going to re-export the bezzle, but actually I don’t necessarily believe that the Asian economies involved ever made any plans based on the assumption that they were going to get paid back full real-terms value. And we’ve got a lot of fixed-nominal-terms debt claims hanging around the show, and everything feels like hell, and the main reason is that all the debt is still hanging around, but the wealth has gone, and everyone has realised that so far from being at their ideal consumption/net wealth ratio, they are miles and miles from it. (This, by the way, is my response to Brad DeLong; we’ve had wealth-destroying events in the past, but fifteen to twenty years into the Great Trans-Pacific Imbalance, there is just sooooo much more debt than there used to be.).
The actual value of housing assets has been falling since 2006; the ability of the household sector in the USA to continue to deny this (and expand the bezzle – it’s perfectly theoretically possible for the value of housing to fall as its fictitious component continues to expand) has been gone since 2007. The Greek and peripheral European sovereigns’ ability to roll over the EMU Put has been gone for a couple of years. The demand is gone, and therefore so is an equivalent proportion of investment, and so on; when the bezzle goes, it has a multiplier on it like any other exogenous shock. All that can be done is, in the immediate term, to replace as much of the lost demand as you can with government demand (complicated, obviously, within Europe), and in the medium term to use as much inflation as the grey gnomes will let you get away with to close the gap between nominal debt and nominal income/wealth in the sensible, painless way.
But what of the banks? Well, what of them? You’ll notice that they don’t have much agency in this model. That’s because I don’t think they have much agency in real life. The Enron/Worldcom/GX bubblet probably marked the chilly limit of their ability to push the credit bubble. If the banks really could systematically supply-push, then there would be a credit channel of monetary policy and some empirical confirmation of Bernanke & Blinder (1988). But there isn’t. In fact, the banks aren’t lending, not out of lack of confidence or ability to trust one another – they’re lending for the more immediate reason that there is nothing out there to lend to. Which suggests to me that a) there is no particular reason to stuff them full of liquidity in the hope that this will “get them lending again”, rather than building up risk-free profits and not doing anything with them, and b) there was not necessarily all that much point in bailing them out in the first place. The Credit Anstalt failure triggered the 1930s depression, but that was not today (“I saw Montagu Norman in his prime / Another time, another time“). Specifically, back in the day, bank failures led to wealth shocks (and deflation led to real wealth shocks) because there was a gold standard and no deposit insurance. These days, not so much.
ENVOI: Very bad things happened during the run-up to the crisis. Things were done which were dishonest; things were done which even to this day probably make it administratively more inconvenient to work out the process (although really, and this is my reply to Alex Harrowell part 1, foreclosures are very much a second-order issue. It is terrible to be foreclosed and even more terrible to be foreclosed in a stupidly-organised process than a good one, but we are not in a depression because of foreclosures; we’re in a depression because of the massive wipeout of equity in houses of people who are still current on mortgages and always will be).
Really, though, equally bad things are done every year, boom or bust. 2004 was a really great year in the financial markets, apart from for the people who were mis-sold payment protection insurance by the UK banks that year. 1996 was also a good year, unless you happened to be separated from your pension by the “income drawdown” fun and games. All through the boom years of the dot com boom, Household International engaged in absurdly aggressive marketing practices, aimed at some of the poorest people in America, with utterly usurious interest rates widely documented by the past tense community operation ACORN. All these things matter, and I’m wholly in favour of the most condign punishment of those responsible for them. Not least, getting dishonest people out of the industry frees up market share for me.
But it isn’t as procyclical as you might think. All the lurid stories about sales forces on methamphetamine, forgery of documents and so on – I’m sure they happened, and it’s surely a good use of someone’s time to prosecute them. But they’re second order. They really are. At the end of the day, the economic function of the financial services industry is to reconcile the desire of savers to have instant access to their money, with the need to invest that money in projects that only pay back over time. In order to do that, you have to set a price to get people’s money into and out of their investments, and the process by which that amazingly complicated social reality is organised is 1) much more difficult an administrative task than you’d think, as evidenced by the near-total failure of all efforts to put it on a rational centrally planned basis (this, at base is why it’s so wonderfully well-paid), and 2) totally dependent on thick social institutions of trust and good faith. And because of 2), it’s a uniquely fertile environment for the kind of person who preys on thick institutions of trust and good faith. That’s the majority of the basis for regulation.
But the rest of the basis for regulation – it’s to ensure that the system functions in a predictable way. The whole basis of monetary policy presumes a reasonably constant capital/assets ratio in the banking system, and a reasonably constant response of credit to money. There isn’t a separate credit channel, but the lending market is a market, and it has to be possible to make it clear at a price equal to the policy rate. If you have wild swings in capital ratios; if you have the kind of agency in the financial sector as a whole that people want to use as a rationale for blaming the financial sector as a whole for the depression, then the financial sector is making monetary policy on its own, as I said above. That doesn’t seem like a good idea to me.
Higher capital ratios reduce the gearing of the system. They don’t make any difference in steady state, but they make it more difficult to expand credit during booms and they mean that capital is freed up more quickly when balance sheets shrink. There’s a tradeoff here between agility of the financial sector to finance genuine economic expansions (or indeed to intermediate policy-caused imbalances), versus the susceptibility of the system to asset bubbles, and there’s a sensible debate to be had here. But it doesn’t make sense to believe that any regulatory tweaks or changes are going to make the banking system into a new surrogate monetary authority or to delegate cyclical policy to it, let alone currency policy, or even foreign policy (since the accomodation of the trans-Pacific trade was to substantial extent a geopolitical decision), particularly if you’re going to ask them to take on this new responsibility for half the wages.