At last, a brief article on the financial crisis that goes behind the facade to look at some of the deeper structural issues.
The author is Satyajit Das and the article ("Built to Fail ") was published in the latest Monthly. He sees the principle cause as excessive debt:
The most important lesson of the financial crisis may be that the current economic order was built to fail, for the global economy used debt and financial engineering to enhance growth, requiring ever more stimulus to maintain performance. The spike in debt globally caused a spike in growth rates. As much as $5 of debt was required to create $1 of growth. Approximately half the recorded of growth in the US over recent years was driven by borrowing against the rising value of houses (that is, mortgage-equity withdrawals). As the level of debt in the global economy decreases, attainable growth levels also decline.
This is now a fairly widely held view. More interesting, I think, is the beautifully simple fashion in which he goes on to consider the real-world impact:
The world economy used debt to accelerate consumption. Spending that would normally have taken place over many years was squeezed into a relatively short period because of the availability of cheap borrowings. Business over-invested, misreading demand and assuming that exaggerated growth would continue indefinitely, creating significant over-capacity in many sectors.
Until household balance sheets are restored, a significant portion of demand is quite simply gone and hence the capacity created to meet it is of questionable value. It's a brute fact that can't be easily papered over; it will take time, and considerable pain, to make the necessary adjustments. It's also why all the frantic efforts to reflate, to get the credit machinery running again, to encourage consumption, may more often than not fail to gain purchase:
The current initiatives of governments and central banks are a hair-of-the-dog treatment. The problems they seek to address can be traced to the high levels of debt accumulated by banks, companies and consumers. In effect, this is now being replaced by government debt and, simultaneously, the debt-fueled consumption of companies and consumers is being replaced by debt-funded government expenditure. Yet adjustment in the level of debt and asset prices is part of the process through which the global economic system will re-establish itself. Like King Canute, central bankers and finance ministers cannot hold back the tide.
Nevertheless, they all feel they must try and already the severity of the crisis has created a "Whatever It Takes" attitude. This is (arguably) all very well if it works, but if it doesn't it risks utter disaster. Not only economically, but in terms of social stability and trust in the political system.
We better hope they're right.
I think "utter disaster" is over the top. If we're bogged down in bad debt for years (like Japan) or have higher levels of inflation for a reasonable amount of time (like the 70s), it might not be the land of milk an honey, but it's hardly utter disaster.
Hey Conrad,
Obviously you have bought your plasma TV. I haven't. It is an utter disaster!
If they're the two worst alternatives, Conrad, then you're right. I hope that's so but fear it's not. Neither the stagflationary 70s or Japan's malaise were of the same order of magnitude as this crisis.
As just one small illustration, consider US debt ratios in 1975 compared to now. Then, overall credit market debt outstanding was 160% of GDP; today, it's 370%. Domestic non-financial sector debt was 138%, now 236%. And the killer, financial sector debt then was 16%, now 121%.
Lest we take too much comfort here in Australia, our debt/GDP ratios are in many cases as bad, and in some (like household debt) considerably worse.
Perhaps in the end this won't matter as much as I think it will and all this debt has only a tangential relationship to reality. If so, we needn't fear catastrophe, or anything like it. Truth is, I don't think anyone has definitive answers to these questions; over the last decade or two we've voyaged a very long way into terra incognita. Any historical analogies are approximations at best and, as has become only too obvious, economists are for the most part utterly adrift.
I guess the one thing we can be certain of is that it won't be boring.
Whilst not questioning the general validity of the premise, I do wonder about the "too much debt", often vulgarised into "living beyond our means" in the context of a global recession.
After all, the net debt of the world economy is 0, just as it had always been, unless Roswell's visitors were financiers or the worst nightmares of David Icke are true.
I guess this can be clarified into "too much poor debt", but that ends up being more a regulatory problem, and less a morality tale.
(This isn't about Das' article specifically though I should add)
Perhaps we have different definitions Ingolf. When I think of "utter disaster", places like Sudan come to mind, where it really is an utter disaster. I can't see how Australia would get even close to anything like that.
I guess an alternative way to look at it is to try and predict the worst possible scenario that could eventuate. On this note, let's say that the Australian dollar falls to 25c to the US dollar and we get an 18% unemployment rate. That would surely be bad, and we'd have a huge cut in living standards. However, it still wouldn't be the end of the world -- we'd have to save for a few months to buy consumer goods like we did for most of the 20th century, people would have to start riding scooters about like most of the world, and the actual percentage of the population employed wouldn't be too different to a lot of the 20th century either (it's just there would be more unemployed males). That's certainly bad compared to now, but, at least in my books, it isn't an utter disaster.
Although the whole thing can be made into a morality tale, Richard, that certainly wasn't what I intended.
As you nicely illustrate, it does all net out, but the individual debts are no less real. Also, whenever debt growth consistently exceeds economic growth, a lot of "poor debt" will be created. Those who borrowed in order to buy now rather than later won't be able to buy later until their finances are back in shape. At an individual level, this is neither here nor there; it's simply a matter of personal choice. If the habit becomes sufficiently widespread, however, it will eventually bring on a recession or even, if the extremes are great enough, a depression. Somewhat simplistically, that's what I think has happened.
I think you're dead right that it's principally a regulatory problem, although in my view it's also a monetary problem. Still, in many ways these are merely different aspects of the same deeper issue. Namely that if financial heart attacks are to be avoided, ways must be found to effectively restrict leverage and limit credit growth. The great deregulation of the 80s and 90s (while in many respects valuable and necessary) also unleashed a monster; unfortunately, this wasn't widely understood. Instead of finding ways to keep it reined in, to maintain prudent capital ratios and to ensure transparency, the financial sector was, particularly in countries like the US, set free to run amok. That it was able to run so far astray was in turn, I believe, largely the fault of a monetary regime which accommodated the growing excess, sometimes willingly, sometimes less so.
Fair enough, Conrad. In those terms, I entirely agree with you.
Hopefully I can recycle http://mises.org/story/3194 from a few months back:
And in reply to Richard Green:
A simple example: you come to my place once a week to swap tomatoes for a pig, we both get something useful, transactions are completed instantaneously and no debt ever exists. One day you bring your tomatoes and I take the tomatoes but I don't have a pig, "Sorry, big pig next week," I say in the efficient manner spoken by the man on the land. Now you go home empty handed, and a debt has come into the world. There are no new real pigs in the world, but there is one extra imaginary pig that you believe is coming to you.
Should I come good with the promise, the debt is cleared and we are all happy. Should I fail to deliver on time I can either promise more (and hope for a patient audience) or we start a longstanding neighbourly Feud and teach our sons to hate. The US economy has been exchanging promises for larger promises for many years and the audience just got impatient.
I would say, in colloquial terms, a lot of people felt they were promised something and now they discover that the promise ain't gonna happen. Is it immoral to make a promise that you can't keep? Maybe it was all just an unfortunate misunderstanding. My point is that the total debt in the world can be non-zero, and it is all in the form of imaginary goods that people believe they will get at a future date.
As a regulator, the ideal thing would be to make it impossible for anyone to promise anything unless sure that the promise will be kept, but we all know this simply cannot be achieved. All debt comes with some component of risk because we cannot fully control the future. The next best thing the regulator can do is ensure that the person taking the risk understands the true nature of this risk and has a fair assessment of what that is. I suspect that this is also not achievable, but in this particular case the regulators made not the slightest effort to inform anyone which is a morality tale in itself.
Once the promise has been broken, the regulator can give the promise-breaker a good old fashioned country thrashing; which won't create the missing goods, but it might serve to encourage caution next time round. The trouble in this case is the highly interwoven nature of the transactions makes it difficult to figure out exactly who was the promise-breaker (but most people have their eyes on the banks).
Ingolf, I agree that there is a lot of monetary blame to lay. From one perspective even that is a regulatory problem as well. I do agree there is a greater regulatory problem. For one thing, a great many of these 'toxic' assets are really regulatory arbitrage, and that this went on to such a scale without the regulators realising that the rules were being leapt through like fountains on spring break suggets a pretty big problem. Warren Buffet famously made the point with respect to Fannie Mae and Freddie Mac:
(and repeated it, p17)
~ ~ ~
Tel, at least you should be happy that I think the US officially abandoned the fractional reserve banking you are complaining of sometime last year (I doubt it has resumed since).
More to the point, are we sure that promises are being broken? I would have thought not many. Rather, you promised Dick, not a pig, but pig as described in this 50 page document. Dick collected 1000 such promises and promised Jane a particular set of income streams and risks from those 1000 promises, as described in this 300 page document. Jane then used that promise as collateral to support her promise to Harry.
When someone twigged that Dick's pigs were rather malnourished, and might not contain as much bacon as thought, a cascading stream of devaluations went through the chain of promises. So Dick's promise to Jane was now worth less than Harry had thought, so Harry was entitled to ask for more collateral to support her promise. But Harry wouldn't accept such vague promises this time, so Jane had to use other, more traditional, promises to support her promise to Harry.
And her promise from Dick became 'toxic', because it was no longer acceptable collateral at any ordinary rate. So in effect her net stock of promises was sharply reduced, and thus her ability to make credible promises to people.
At this point the government stepped in, assumed a massive promise on behalf of our children to pay taxes, and gave Jane a big bundle of very simple promises so that she could start making promises again.
Or so I understand it, and I apologise for the length, but the point is that the pig is still there, but we know realise that we didn't really establish what kind of a pig it was and how much lipstick it was wearing. The promise wasn't broken, we just didn't really know what it was.
(And then, yes, you were sacked, and you did break your promise to Dick, and you are now bankrupt with no assets, and that cascades through the system too, but this is really as much an effect as cause).
To (finally) make my point, which I hope marries your concerns with Ingolf's, I will paracite JKG (I can't find the original, maybe I am making it up), whom I dislike and generally disagree with but has a point on this:
Yes the story does sound like the Austrian heresy. ;-) Still, Richard's point is correct in that each loan has a lender and a borrower, and tel, lenders don't spend money they have lent to other people so you fractional reserve extra dollars are not all down at Harvey Norman at the same time. As Keating pointed out, consumption in the US and other places (like us) has been partly funded by the Chinese not consuming, so the Chinese (and others) have been under consuming. The problem is not overconsumption generally, but a bubble created by too much lending by the Chinese. CDOs and CDSs just made it a bigger knot and the fricitional damage will be worse as a consequence.
What Das is suggesting is a "do nothing policy". In my view (as I argued in Response to Turnbull, that it is the surest way to build up government debt. It would create so much unemployment and a fall in profits that public debt would have to rise.
Indeed we would end up with a fall in economic growth and a big rise in government debt.
So what he has achieved?
Pedro, I don't see how CDS made anything worse at all. CDOs perhaps yes, because they are the kind of generic instrument I described. But if Jane had bought protection on Dick's promises with a CDS, then most likely, Jane would have been better off. At that point it really would have come down to collateral, and whether Jane, who didn't have enough alternative promises to wholly replace the CDOs (multi-packed promises) she was holding, might have had enough alternative assets to post as additional collateral on the CDOs as they downgraded.
So CDS might not have saved her, but they would have given her a much better shot. They certainly wouldn't have made it worse.
Fred, I realise that you know a lot more than me about this. But are you sure that a do nothing policy even could build up as much debt as the US seems hell-bent on taking on? Or is your answer that they will get back a lot of their money so the debt isn't as bad as it seems?
Patrick,
I know for sure that public debt will need to increase as economic activity declines. What I don't know is how much US pump priming will increase debt, but it will certainly be at a lower level of economic activity, even if it is not proportional.
I do happen to believe that the large government debt will take a few years to run off but it will do alot of good in the meantime.
Fred I can agree in the Australian context, I have no problem with the Gruen/Argy public-debt-is-not-as-bad-as-it-is-made-out-to-be-at-least-in-Australia school, indeed I think it is almost common sense, but I am interested in whether you think this is applicable to the US where the figures are such an order of magnitude higher at the best of times and are a couple orders of magnitude higher now, even expressed relative to GDP.
I would prefer that we factor in a smashingly conservative $15trillion in unfunded medicare and federal pension entitlements (I have seen 36tr touted for Medicare alone) and that is ignoring the States. With US GDP of 13 trillion and Aust of 1, we would have to be borrowing 2 trillion to catch up with Medicare, TARP, and all the rest.
Do you think that would be sensible? Do you think it is sensible for the US, or do you think that the US is sufficiently 'different'?
Don't know for sure. I would hate to see it decline sharply, as it is an economy which once it takes of, is likely to rebound quickly.
Patrick, I think CDSs contribute to the problem by the sense of security they provide, thus increasing the appetite for more of the problem debt.
I think govt spending now (and thus debt) is only justifiable to the extent is eases the adjustment process that Das is talking about. For example, I can't see much point in supporting jobs in consumer retail because we almost certainly face a contracted retail sector for years. But the do nothing school has to answer the objection that doing nothing will lead to a big overshoot downwards as each stage of the developing recession causes new problems.
Patrick, agreed. Figuring out the dividing line between structural and regulatory problems when it comes to money and credit is never going to be easy. They'll always overlap, and in any case much of it probably comes down to semantics.
Fractional reserve banking hasn't gone away, the machinery is just constipated. Banks are undercapitalised and hence severely constrained in their capacity to lend; in addition, and perhaps more importantly, there aren't a lot of keen borrowers anymore and of those that can be found, there probably aren't that many you'd really want to lend to. As a consequence, commercial bank loans and leases have actually declined by 7.5% annualised over the last three months and by a little over 15% annualised over the last month. The only real growth area on their balance sheets has been cash, which has more than tripled in the last year (from around $300 billion to over $1 trillion) as the Fed injected reserves. Based on recent policy announcements, it's going to grow a whole lot more.
I have mixed feelings about your quote from Galbraith; "trust, but verify" feels more right to me. One of the great flaws in the current structure is that it fostered a widespread belief that the authorities were doing the verifying without actually delivering on the implicit promise. This (along with a great many other things, such as the role of the rating agencies) made it all too easy for participants to close their eyes to the risks they were in fact taking.
On the derivatives issue, I'm inclined to side with pedro. Although in themselves they're a zero-sum game (and as you say can be very useful for managing risks) I do think they facilitated the explosion of credit. Had they been exchange traded and cleared, I don't think they'd have been a problem, but unfortunately the US made the fatal error of leaving them OTC and almost entirely unregulated. As a result, poorly collateralised (or entirely un-collateralised) contracts proliferated like confetti.
Fred, I don't know that Das is suggesting a "do nothing policy". For example, he says at one point (when discussing the failure of policy actions to achieve traction) that "Well-intentioned infrastructure spending will take some time to have any meaningful effect." I thought his piece was more directed towards gaining a structural understanding of why (and how) things ran off the cliff, because without that understanding, policy formulation will be operating blind.
Pedro's formulation in #15 struck me as pretty good; "I think govt spending now (and thus debt) is only justifiable to the extent it eases the adjustment process that Das is talking about." The greatest danger, I think, might stem from a failure to accept that debt levels must be allowed to come down. Not only is this likely to produce policies that impede the adjustment process, it could also (in some countries like the US), actually place the financial system itself at risk.
I liked this paper on CDS, it seemed consistent with my expectations and understanding. It attemps to address, in particular, your complaint about the OTC market.
Where we differ is that you are probably hopeful that new regulation will be positive (for example you are a fan of banning OTC CDS) whereas I am certain that the costs will outweigh the benefits.
But I also do agree with Pedro's statement about adjustment. My mind literally boggles at the money given to GM. Imagine if that money had been divided amongst the workers, even including those of the supplier networks, as retraining and relocation subsidies. Anyone who believes that the money given to GM/Chrysler was a loan should be paying face value for sub-prime RMBS.
Patrick, much of my concern stems from deposit taking institutions (and others deemed "too large to fail") being so heavily involved in derivative markets (and in dealing more generally). It would, I think, be far more sensible if they were precluded from doing so. Strict hedging perhaps, but nothing else.
Were that the case, whatever virtue I see in closer regulation elsewhere (much less in the banning of contracts) would diminish rapidly. There would still be benefits in mandating a reasonable degree of transparency (and perhaps the collection and collation of up-to-date data) but beyond that, providing it was made crystal clear to all participants that the cavalry wouldn't arrive if things went badly wrong, I'd be happy for that part of the market to operate very much on a caveat emptor basis.
Still, this is all edging more towards the large (and fractious!) topic of how the financial system more generally should be structured. By the way, thanks for the link to the article. I've only glanced over it, but it seemed quite good.
I am with Patrick in fearing the new round of regulation. My view after many years as a lawyer is that regulation to protect people from their own stupidity has high and unforeseen costs and little effect. I often wonder if the existence of ASICs and so forth make the problem worse by disguising that fact that securities investment can be a leap into a bear-pit.
If we accept that too big to fail is a political necessity then perhaps the best regulation is to limit the number of too-bigs? I really don't like mucking around with the market, but it is impractical to pretend that moral hazard problems can be avoided by political will, so we only have the regulatory environment to help in that respect.
I have always like the 4 pillars policy because overly large banks are a systemic danger.
We should not forget that the greatest cause of the present problem is the amount of money being lent and the low interest rates that encouraged the borrowing.
Be careful Pedro you will be a free banker soon!
Ingolf, you have hit a real issue. What did GE, AIG, Citibank and RBS all have in common? They leveraged the stability, cashflows and, crucially, credit rating, of one business (diversified manufacturing, insurance, deposits and mortgages) to give them a competitive advantage in another business (financial products/speculation).
I don't even know if this is good or bad. Bear Stearns and Merrill Lynch went arse-up without any such balance sheet arbitrage. Goldman Sachs and even eg Macquarie are still going also. But I can't help but wonder if there is a problem created when one entity can price an identical transaction differently because of their related entities' credit ratings. Do credit rating agencies have any real idea what kind and size of implicit call is extant on a given company's rating?
Patrick, I think free banking is one of those ideas where it sounds good at first, but nagging worries set in and you can't quite sell yourself on the idea. I think maybe some things have moved on since the 19th century.
Arnold Kling on the financial regulation chess game:
http://www.finreg21.com/lombard-street/the-chess-game-financial-regulation
"Regulatory systems break down because the financial sector is dynamic. Financial institutions seek to maximize returns on investment, subject to regulatory constraints. As time goes on, they develop techniques and innovations that produce greater returns but which can also undermine the intent of the regulations."
and
"Instead of trying to make the regulatory system harder to break, we might think in terms of making it easier to fix. "
and
"The best way to make our financial system easier to fix would be to reduce the incentives for high leverage. We promote home ownership by subsidizing mortgage indebtedness. It would be better to provide subsidies and encouragement toward saving for a reasonable down payment. Likewise, in the corporate sector, our tax structure tends to penalize equity finance and to reward debt finance. Changing the system to tilt more in the direction of equity finance would go a long way toward reducing the vulnerability of our economy to crises at banks, insurance companies, and investment banks."
"Do credit rating agencies have any real idea what kind and size of implicit call is extant on a given companys rating?"
Reliance on ratings agencies is the same problem as that I alluded to in comment 19. And do we detect the protective hand of government in the glorification of the ratings agencies?
It's easy to be an anti-government cynic when, you know, there's so much to be cynical about.
Yes, pedro, I think there's every reason to fear that many, if not most, regulations will have unintended effects. From your comment, you've plenty of practical experience of this unhappy truth. Still, as you say, the nature of our political system makes any move back towards laissez-faire conditions exceptionally difficult. While such an approach might often be a case of "being cruel to be kind", it's hard to see where the constituency for adopting it might come from. That being so, I guess we're stuck with regulations and various safety nets.
This needn't be all bad; indeed it's not too hard to imagine structures that might work quite well. One example would be some variant of narrow banking (which would be exceptionally tightly regulated and guaranteed) to handle the system's need for on-call money, with all other institutions having to do with savings and investment forbidden from taking deposits and non-guaranteed, but otherwise allowed to ply their trade governed only by the common law and some minimum amount of specific regulation. Not only would such a structure ensure that most longer term funding was (in my view properly) viewed by investors providing it either as equity or a longer-term fixed interest commitment (with all the systemic stability that would follow), it would also slow, or perhaps even halt, the relentless depreciation to which fiat currencies have always been prone.
Your quotes from Kling suggests he's keen on heading in a similar direction, albeit in a different fashion. His comments made a lot of sense. As for your #22, I think we do, and yes there is.
It's an interesting question, Patrick. I don't think it need be a problem except when the operation free riding on a "parent's" credit rating is either dealing in a sufficiently arcane field to be in effect freed to do their own thing, or of course if it manages to slip the leash of parental control. AIGFP was arguably a classic example of both problems.
Steve Keen (and maybe other people) have explained the problem with the CDOs in attractively simple terms. It goes a bit like this:
If I have a single investment, with some smallish percentage chance of failure then I stand to potentially lose the lot. If I have a great many small investments, each with the same smallish chance of failure then the chance of losing the lot is vanishingly small, because multiplying all the small probabilities together results in a rapidly shrinking chance of total failure. Right?
Wrong, actually.
The statistical analysis that was used to get estimated for CDOs was based on a presumption of non-correlated failure modes (i.e. each investment was an unrelated, independent random variable). However, what actually happened was that default on housing payments (and credit cards, and similar loans) turned out to be highly correlated to a downturn in the market (and worse, when the defaults added up they caused further downturn). The high correlation of these "random" failures implied that the CDO was absolutely no protection at all.
In terms of the larger market stability, the CDS system has exactly the same problem as the CDOs but it is more difficult to visualise. Let's consider that a CDS is really just an insurance contract, and the insurance company maintains a certain amount of liquidity to ensure they have enough to cover the random fluctuations in the risk that they are underwriting. The insurance company is quite safe so long as random events happen as independent random variables, they are foobar in the face of highly correlated failure modes (i.e. a market downturn). When the insurance company runs out of liquidity is exactly when the CDS is most needed and the buyer finds that the insurance company can't keep their promise. Thus the CDS is also absolutely no protection at all.
Should the people calculating the risk have been aware that correlation was going to screw with their results? In the cold light of hindsight (and having it explained to me), it does seem bleedingly obvious. I'll be quite honest and say that I doubt I would have spotted the problem in the heat of a boom time, but then again I'm not paid even a fraction of the income of the people who were paid to check this stuff out, so presumably they earned that money doing something?
Ingolf, I'm not suggesting a move to laissez-fair, just a lot more deliberation and humility in making regulations.
No, I understood that, pedro, and entirely agree there's a need for changed attitudes.
Tel, I think you're right that they should have been aware. Unfortunately, the whole financial system was structurally geared towards growth and short-term profits, and this was backed up by the prevailing consensus. Given our generally tribal natures, very few people are willing to leave the herd and venture out into the cold alone.
Patrick, you're probably well aware of the "Big Bang" protocol that's just been introduced by the ISDA. It seeks to address some of the concerns we briefly discussed earlier:
http://www.ft.com/cms/s/0/a645fa74-239c-11de-996a-00144feabdc0.html
The industry has also been working hard at isolating offsetting contracts and "tearing them up". By doing so, it's apparently managed to roughly half the nominal outstanding to under $30,000 billion.
Anyway, these appear to be very constructive steps.
Very constructive steps, and all with out regulatory help from Big Brother it seems. Funny that. If I remember correctly, the market had called out Enron well before the regulators got to the scene of the crime.