Robert J. Shiller, The New Financial Order: Risk in the 21st Century
(Attention conservation notice: approximately 1600 words about a book that I hated hated hated. Shiller’s heart is in the right place, but his mind is somewhere in the gamma quadrant.)
I’d half-expect Robert Shiller to be cursing his luck. He writes a book in 2003 singing the praises of the finance industry — a book that could easily be read as a paean to securitization and the efficient-markets hypothesis — and 6 years later that industry is about as reputable as phrenology. I only half-expect that, though, because Shiller may well be insane. This is an insane book. Maybe “visionary” or “idealistic” is the way we pronounce “insane” in polite society.
His heart is in the right place, which is why it pains me somewhat to say that about it. First, then, let’s take some time to give credit where credit is due.
Shiller wants to insure people against risk. Not just a little bit of risk here and there, but quite massive risks. Right now, we can insure ourselves against our homes burning down, but we can’t insure those homes against a fall in their values. We can insure against our getting sick, but coverage in the U.S. is spotty: it often won’t protect you if you have a pre-existing condition, it’s subject to adverse selection (only the sickest get insured) if people aren’t compelled to get it, and so forth. We collect unemployment insurance if we lose a job, but we’re not protected against the much larger systemic risk that our entire industry will disappear and we’ll be stuck with skills that no one needs. (Think of those who operate printing presses for newspapers.)
The largest risks in our lives, and the largest risks to our society as a whole, are not insurable. Shiller wants to remedy that. He wants insurance companies to write policies against declines in the value of your house. He wants aspiring biologists, at the start of their careers, to be insured against the disappearance of their particular specialty.
Stated at maybe the broadest level, he wants society to share many more risks than it does now; Jacob Hacker, coming at this from an entirely different angle, called for us to realize something similar: that we’re all in this together, and that we have a stake in seeing our brethren sleep more easily at night, knowing they’re protected.
Shiller would scale out even further: he wants, in essence, every human being on earth to protect each other. Poor countries would enter into swaps with rich countries, trading one income stream for another: if the poor country’s GDP grew faster than expected, it would owe money to the rich country; if it grew more slowly than expected, it would receive money from the rich country. (If both countries’ GDPs dropped by the same rate, neither country would owe the other.)
Here’s where most readers are going to howl in protest, and rightly so. Let’s imagine the two swapping countries are Nigeria and the United States. How, exactly, will that work? Do we really trust the Nigerian government to share windfall GDP with us? Of course not. Shiller doesn’t even address this thorny issue. The closest he comes is a swap between India and the United States. India, he says, will surely pay us off; the historical evidence suggests that it, or at any rate developing countries like it, have a respect for contract. Even the new World War I-era Soviet Union honored czarist debts, says Shiller. But we have reason to be more skeptical about Zimbabwe than about India. Shiller obviously realizes this, yet he doesn’t mention it. His only examples are comparatively easy cases. What about Burma? Should the U.S. government be any more willing to enter into a swap with a repressive regime than private investors are?
But again, his heart is in the right place. That phrase looped over and over through my head as I read The New Financial Order; it was the only way I could make it through. He wants us to be our brothers’ keeper, no matter where our brothers live.
Not only is the problem that we each suffer too much risk; it’s that we don’t take the right kinds of risk. Maybe we’d be much happier as painters or teachers, but there’s too much risk in those jobs. We go on to get business degrees, spend our lives in a cubicle, and die unhappy. Shiller wants us to be able to take those risks, and he wants insurance companies to help. He gives the example of a student deciding whether to pursue a career in recombinant-DNA technology. The risks are high and the rewards may be great. An insurer should be willing to exchange in a swap on the student’s future income prospects, insulating the student on the downside and profiting on the upside.
Any such insurance policy has to face the moral-hazard risk: if I’m insulated from downward movements in my income, I have less incentive to work hard. The insurer may, then, only insure against a downturn in my industry. That is, if recombinant-DNA research goes the way of the newspaper industry, I collect from the insurer, but not if I lose my job when the industry’s doing fine. Let’s divide up these two sources of risk: call them “macro risk” and “personal risk.”
Shiller elaborates the macro side more than the personal side. Policies based around macro risk are less subject to moral-hazard concerns: it’s assumed that I will have very little control over industries as a whole, hence have no effect on macro risk. Shiller never asks how much of my risk is macro and how much is personal. That is, even if an insurer could protect against macro risk, how much would that matter to me?
Here’s where The New Financial Order tips over from merely dreamy into straightforwardly batshit insane. Both to reduce moral hazard, and to better quantify macro risk, Shiller is pushing for a massive intrusion into our personal lives that he called a GRID: the Global Risk Information Database. It’s a worldwide computer system, presumably run by (does he ever really clarify who runs it?) insurers and banks. It will track as much about us as is necessary to help our insurers. And of course it will be carefully guarded; the insurers will likely hire T.J. Maxx’s crack team of security gurus to guard the GRID.
The GRID takes on more responsibilities than just risk analysis, even within Shiller’s book; I shiver when I contemplate what the GRID would do in the real world. By page 224, it’s helped improve on the Jobs Rated Almanac (which classifies jobs by how enjoyable they are, how risky, and so forth); by the next page, Shiller tells us that the GRID will contain information “not only about individual incomes but also on other individual circumstances that could someday be related to terms of risk management contracts.” What, pray tell, could not someday be related to terms of risk management contracts? I don’t think I’m being unfair at all if I suggest that Shiller would like to see us all equipped with probes that check whether we’re having unsafe sex or drinking too much coffee.
The overarching problem is that Shiller is a technocrat — a “high modernist,” to borrow James Scott’s terminology from Seeing Like A State. If only the world can be made to fit the technocrat’s plan, all will be well. If only Bob Shiller, with the aid of the insurers, is allowed to rationalize the world, the existing world’s chaos can be brushed aside.
Bob Shiller will bring to this massive undertaking his expertise in computer science:
A GRID could also be designed to provide an infrastructure to faciliate the handling of last wills and testaments. The system that we have is expensive, relying on lawyers and trustees. Only the wealthy can easily make complicated plans for the use of their life savings. Moreover people have no effective way to publicize the charitable parts of their wills if they want to. For these reasons, perhaps, most people leave nothing to charity in their estates … A pleasant, user-friendly interface on the GRID to design one’s will, allocate some of the estate to genuinely low-income people or to other causes, and publicize that one has done so, might increase the amount of such giving and make it more effective.
Yes, the reason people don’t give away a lot of money at their deaths is that they don’t have a pleasant UI to do so. They would cut their children out of their wills and donate directly to the poor if only they could. The GRID would help with this because … well, it just would. It’s a computer, all right? It’ll do what computers do. It’ll have buttons and stuff. You can use a mouse.
Among the living, insurers are gathering up risk from all of us little people. They’re packaging it together, building a diversified portfolio out of it, and making tons of money. Everyone wins: they get wealthy and we breathe more easily. Let’s gratefully turn our eyes up to Shiller to ask who these heroes of capitalism are:
Dramatic innovation has … come from investment banking firms such as Bank of America, Barclays, Bear Stearns, Citigroup, Deutsche Bank, Goldman Sachs, Hongkong and Shanghai Banking Corporation, JP Morgan Chase, Merrill Lynch, Morgan Stanley, Société Générale Group, and Wasserstein Perella. More innovation has come from insurance and reinsurance companies such as ACE Group, Aegon Insurance Group, AIG … from mortgage and consumer finance firms such as Fannie Mae, Freedie Mac, and GE Capital; …
Truly an honor roll in which only the best and brightest are allowed to stand, backs straight, benign visages fixed heavenward.
Gaze upon them raptly, for within them burns a gentle flame that shall light all mankind. They are our saviors, and Bob Shiller is their prophet.
So, one thing I don’t get about wide-spread insurance is that it’s not clear that the insurance company model even works for large-scale insurance. Insurance policies are only worth anything when the insurance company can follow-through and it seems that there are situations that seem to happen, relatively often, where insurance companies are nearly wiped out by certain big events (big hurricanes, financial crashes, etc.). Basically, if large proportions of people are getting screwed and income streams dry up, then, presumably, the income streams that insurance companies have dries up as well.
Now, I get that insurance companies are motivated to have diversified portfolios, but when times are really hard, even diversified portfolios can lose substantial values and that’s also coincidentally the time when people are also looking for insurance payouts.
Basically, it seems to me that insurance companies really can only provide safety nets up to some catastrophy threshold. Governments, on the other hand, seem to be able to provide a safety net at a considerably higher catastrophy threshold…though even governments have their limits.
Basically, if individuals need help in contrast to a stable background insurance company protection works. If everybody’s getting hit, insurance companies aren’t really as useful.
So it seems silly on it’s face for everybody to get in bed together to try and insure each other. it just doesn’t work. How is it that people missed this? Were they all just assuming a stable background?
Comment by mrz — March 29, 2009 @ 10:53 am
What you mean by “stable background” here seems to be “when the destruction isn’t universal.” I think quite often the destruction isn’t universal. Japan was in a recession for a decade, and the money flowed elsewhere. So Mexico had a boom for a while. So did the U.S. We now happen to be in a situation where there’s nowhere for money to go: everyone’s screwed at once. So it would seem hard to insure against the trouble we’re in now. It’s not always going to be so hard. In fact it’s probably rarely going to be so hard. Insurance companies generally do just fine. AIG’s insurance division — not the part that was writing CDSes — is still making money. Real insurance (again, not CDSes) can rely on the Central Limit Theorem, and it has very stable risks. The government can mandate that insurance companies hold a certain amount in reserve for the real catastrophic events (e.g., once-in-a-century typhoons).
Comment by slaniel — March 29, 2009 @ 3:05 pm
I agree with mrz. We recently watched the Spike Lee documentary or Hurricane Katrina, and of course the insurance companies look very bad. But it seems reasonable to me that insurance companies shouldn’t have to deal with massive catastrophes. It would probably be more economically efficient to limit their purpose to individual bad luck, and have all of society (however defined) step in for massive events like Katrina.
Comment by Adam Rosi-Kessel — March 29, 2009 @ 5:09 pm
But isn’t that exactly when you want insurance to function?
Basically, you buy insurance for stuff you don’t expect to happen. What’s the point in having things like insurance against the economy collapsing? If we’re all scratching around in the dirt, the insurance policy from AIG you have to pay out a million dollars to you isn’t going to be worth the paper it’s printed on.
Basically, what I’m saying is, it just seems like some things just aren’t feasibly insurable. This is not to say that insurance is somehow impossible for everything, just that I think extremely large events with catastrophic consequences can sometimes exceed the scope of ability to practically insure. Worse, it would seem that a particular web of unforseen events can sometimes manufacture itself into an extremely large disaster… like the current financial crisis. If some wall street banks take a dive, which leads to credit markets failing, which snowballs into tremendous money loss all around, then where is the insurance company going to come up with the money to pay on its obligations? It doesn’t make sense for a company to hold too much in reserve, otherwise they are just a glorified bank. So it seems there must be some break-even point past-which you can’t reasonably be expected for insurance to come through.
Comment by mrz — March 29, 2009 @ 6:38 pm
Adam: I can see having the government insure against massive catastrophes, and leave private insurance for the smaller stuff. Isn’t that exactly what we do? If an insurance company gets hit with a catastrophically large set of payouts — like from a hurricane — even after they’ve done their due diligence, the government should probably bail them out.
At the same time: the government might well want to require larger reserves from insurers. I wonder, for instance, how likely the insurers estimated Katrina was. We hear about “once per century” hurricanes that seem to happen every quarter-century or so. Maybe the government should apply closer scrutiny to insurers’ risk models. Maybe the insurers ought to be charging higher premiums for homes on the gulf coast. In fact, I imagine that if insurers know they’ll be bailed out in the event of catastrophic damage, they’ll charge the lowest premiums they can get away with.
Comment by slaniel — March 30, 2009 @ 8:34 am
mrz: Yeah, all of that makes sense. Though I think we would need to dig into the details of what is and isn’t feasibly insurable. I dig your point that when there’s a global depression, an insurance policy isn’t worth a thing — in which case the government obviously has to step in and print some money to get things rolling again. So in the event of catastrophic failure — after forcing insurers to perform their due diligence on policyholders — there’s a role that only the government can fill.
But in the event of non-catastrophic failure, are we still prepared to say that certain things aren’t insurable? For instance, is mortgage default generally uninsurable, or only uninsurable in a depression? Seems like the answer here is “no.”
We’re probably saying the same thing; I just wanted to clarify.
Comment by slaniel — March 30, 2009 @ 8:42 am
I’m note sure. I want to say “no” too, but I have a sneaking suspicion that there may be a smaller class of failures for which insurance is also not suitable. So, a single mortgage failure or some small number of failures are insurable. That seems clear. What’s not clear is when we move beyond simple failures into the domain where you can have interconnected failures that ruin the hedging done by a firm’s or group of firms’ portfolios. I’m wondering if there are cases smaller than depression-scale failures that could, say, eliminate the ability of one company to insure. Though, I suppose if that company had insurance from other companies, for it’s failure, that could be covered, but is that really workable? When does a single insurance company fail in such a way that other insurance companies don’t also fail?
Basically, if a depression-type scenario happens because of “cycles” in the insurance graph where I insure you, and you insure me, and we both get hit on both ends, can we construct a situation where you have cycles in subsections of the “insurance graph” that take down the subsection but leave the rest of the graph intact? That is, suppose insurers for a particular thing all get wiped out? I suppose this is just bad planning on their part? They should have sought further insurance?
So, yeah, I’m not sure how to construct a smaller failure scenario than a depression or just crappy planning, which presumably “don’t count” here. I’m not sure what other scenario would exist that would cause a smaller, non-system-wide failure.
So I guess I provisionally agree with you? :)
Comment by mrz — March 30, 2009 @ 10:45 am
BTW, on the problem of hurricane insurance, there’s a bit of a paradox there. If you mandate that insurers have bigger reserves, then they will charge more for insurance. However, you might be surprised to know that hurricane-prone areas like the Gulf Coast actually include a lot of really low-income people. So, faced with higher insurance, many people will opt out. This means a lot more uninsured people to deal with after a hurricane. If you then force municipalities to require insurance this means forcing people to sell their homes and either move away.
So, such a plan would basically displace poor people. I’m not sure how you deal with this other than to say “No poor people on the Gulf coast” or you have to give assistance to people who live in such areas and suck up the fact that we’re going to have to sink money into dealing with the issue every 25 years. Though, I suppose the latter might be mitigated by further assistance, like making sure homes in the area meet hurricane codes, and other such preventative measures.
Comment by mrz — March 30, 2009 @ 10:55 am
Apparently, I do not have permission to edit my own post. So:
sed -e “s/either move away/move away/” last_post
Comment by mrz — March 30, 2009 @ 10:56 am
Right, you’re talking about reinsurers. They insure the insurance companies. What’s nutty to me is that reinsurers often are themselves reinsured. When a reinsurer is itself reinsured, this is called retrocession. (I learned all of this from the Wikipedia entry on reinsurance.) The reinsurers need to be careful not to create cycles in the graph, obviously: “A” insures a mortgage, “B” reinsures “A”, “C” re-reinsures “B”, and D=A. Badness ensues.
Again, the ultimate reinsurer is the government. We want this; we don’t want insurers to be able to fail. But in exchange for this protection, they need to be only ensuring things safely, need to have sufficient capital reserves, etc.
In the case of a difficult-to-ensure thing, like mortgages, I don’t know whether the appropriate answer is to forbid ultimate government reinsurance on anything unstable, like mortgages, or just to set the capital-reserve requirements as high as necessary.
Here’s where I want to dig into the numbers: if an insurance company wanted to write policies to protect against even something as terrible as the mess we’re in now, how large would the premium have to be? If you expect this to happen once every hundred years, the premium might not have to be so large: catastrophic damage once per century doesn’t turn into much per month. However, you have to be really, truly sure that it’ll only happen once per century. If your model is off, your premiums will be set wrong. And we don’t want the government insuring unknown risks.
Or maybe we do, to get back to Adam’s point. Maybe it’s good for society to insure against that kind of risk, and have the government do the insuring. But that’s a separate conversation. I think the reason we have private insurers, rather than just having the government insure everyone directly, is that private insurers can go out of business, and hence have an interest in judging risks properly. Government doesn’t have that self-preservation incentive. So there’s obviously a good argument for having private insurers.
Now, if no private insurer will pick up some risk that society thinks ought to be insured — like mortgage risk — then we ought to ask why, and we ought to ask how to structure government insurance. Maybe look at flood insurance as an example. How does that work?
I don’t know if I’m taking us off-topic. I’m trying not to, honest.
Comment by slaniel — March 30, 2009 @ 10:57 am
On your last post, about hurricane insurance: it may well be the case that we shouldn’t be subsidizing people’s habit of locating in dangerous places. There’s a powerful argument in that direction, obviously: if you can’t bear the risks of living in a given place, then why should society be picking up the tab? There’s no argument, it seems to me, that people have a right to live on the Gulf Coast, whereas there’s an excellent argument that people have a right to health insurance. Subsidizing people’s health insurance makes a bunch more sense to me than subsidizing their choice of housing.
Comment by slaniel — March 30, 2009 @ 11:00 am
Oh, a bit more on retrocession: I have a vague picture that CDS issuers often turned out to be reinsuring themselves. I’d like to understand more about how you prevent this.
In general, I have a vague story forming in my head about where risk goes. Everyone is hedging everything: insurers hedge their risks, reinsurers hedge theirs, etc., etc. The story is supposed to be that I can hedge away as much risk as I want, leaving behind a diversified portfolio with exactly the risk and return that I want. That hasn’t quite worked out as planned.
The vague story says: there’s a certain amount of risk in the market. That fundamental level of risk is not going to go away. It’s a “noise floor,” so to speak. You can push it around — hand it from insurer to reinsurer to portfolio holder to, ultimately, government, but it’ll still be there. So microeconomically — on the level of an individual actor — you may be able to get rid of it, but macroeconomically — the total level of risk in the market — you cannot. It’s simply not the case that all actors can get rid of their risk simultaneously.
I know the counterarguments to this, of course. A company like AIG is supposed to have a massive portfolio of everyone else’s risks. All those other risks are supposed to be uncorrelated. If they’re uncorrelated, then the Central Limit Theorem leaves AIG with a bundle of risk at the end of the year that it can quantify precisely. If those risks are correlated, then we have theorems for that. But again: they didn’t help. Is it just that AIG (or whoever) got the correlations wrong? Is it still possible to construct the magically riskless portfolio?
I don’t know the answer, of course, but this is the direction I want my reading to go in.
Comment by slaniel — March 30, 2009 @ 11:07 am
Actually, I think I misstated it. No one posits that all risk will go away. They do posit that we can construct a riskless portfolio by building synthetic options of various sorts. But the supposition is that those people who can bear less risk (think nearly-retired people) can hand it off to those who can bear more (younger folks), and that ultimately those who are able to bear risk the best are those who have the most diversified portfolios (like the AIGs).
Comment by slaniel — March 30, 2009 @ 11:18 am
If we were using the GRID, you could’ve insured against reading this book. See, Laniel? You are the crazy one.
Comment by chris r — March 30, 2009 @ 3:55 pm
Because they are poor and live there already? Also, people are not rational actors and live somewhere because that’s where their family lived since forever. This means people will live stupid places and if enough people do it, what are you going to do? Just say “Oh well, we told you so. Sorry you’re drowning right now.”? I mean, I guess we basically did that already with Katrina to some degree. It didn’t turn out too well then, either.
Basically, if you’re going to be descent you have to deal with both the guy who’s capable enough to insure himself and help out the guy too poor to insure himself but who has to live where he is anyway because you’re a decent human. The question is, how do you explain that to the guy who is capable enough? Or, perhaps, why should you have to? Is he not a decent human too?
But this is a different topic.
Comment by mrz — March 30, 2009 @ 11:41 pm
A few things. First, I only meant to suggest that we shouldn’t subsidize people’s bad judgment now. At the time of Katrina, I raised the point on this blog somewhere that it’s probably a bad idea for people to live in flood- and hurricane-prone areas, but that the government has been insuring those areas for a long time. It would obviously be disgusting for the government to renege on its promises now. That is: the government must pay up on its existing insurance policies. However, I do think that insurance should be phased out.
Second, as for people not being rational actors: I think you’ll find that if their premiums go up to reflect the actual risk of a flood, they will move away. That makes them rational actors.
I can’t think of any good argument for supporting people’s continuing to live in a place where there’s a substantial risk of their being wiped out, and subsidizing that with insurance policies that don’t properly reflect that risk.
Now, let it be noted at the outset that I don’t know a thing about flood insurance, insurance for homes on the Gulf Coast, etc. I could be totally wrong about it. I fully accept the possibility that someone who knows what he’s talking about could make me look like the fool that I am. But in principle, I don’t see why the government owes people anything just because their ancestors lived there. If it’s so important for these folks to live where their ancestors lived, why shouldn’t they pay a cost that reflects that importance? “Because they’re poor and can’t afford it” isn’t really an answer, because again: I don’t see why there’s any societal interest in keeping people — poor, rich, or otherwise — where their ancestors lived. (Note that there may well be a societal interest in putting poor kids in wealthy towns, if doing so gives the kids access to better schooling. I can see other arguments along those lines. But none of them applies here.)
People are forced to leave places all the time because it’s too expensive there. (Think “gentrification.”) Does society have an interest in letting the poor people who used to live in JP continue to live there?
Comment by slaniel — March 31, 2009 @ 8:47 am
I sort of agree but what I’m getting at is that you may not be aware that people live in these areas that are currently uninsured anyway. They own their homes outright because their granddad worked until the day he died to put down the last dime on the $10,000 mortgage. The remaining generations just continue to live in the same house and pay land tax. Some may not even be aware that they need insurance. So they simply go uncovered. Then you have people who might have some form of insurance who are on the edge. If you make their insurance go up, they may simply not bother having it anymore. The choice for some people isn’t between paying more insurance and moving someplace cheaper, there’s the third choice of doing without and hoping nothing bad happens. So if you bump up insurance rates, instead of having private firms defraying some of the cost, you’re shrinking that and putting more weight on the governemnt.
I suppose the question is: is the cost you lost to people doing without offset by the people who do leave? Also, if people leave, does that make demand go down, which makes houses cheaper, which lowers insurance rates, which makes people think about chancing it and buying a house in such locations? Etc. Etc.
Stupid feedback loops ruin everything! :)
Comment by mrz — March 31, 2009 @ 9:38 am
Another reason to make the Government the “insurer” of events like Katrina is that it was the only entity that really could have prevented it.
Comment by Adam Rosi-Kessel — March 31, 2009 @ 12:14 pm
[...] Robert J.New Financial Order, The: Risk in the 21st Century (finished 27 [...]
Pingback by Stephen Laniel’s Unspecified Bunker » Lists of previously-read books — June 28, 2009 @ 12:19 pm