Can Topology Prevent Another Financial Crash?
nautil.us
nautil.us
He was wrong.
Under the new regulatory regime, a re-run of the 2008 financial crisis would have engendered far less FUD and Lehmans could well have survived. Nobody would have been scared that Lehmans' potential inability would have represented an existential threat to their firm because only the CCPs would have been exposed. (This is a gross simplification but you get the point.)
However, as the article makes clear, the counterparty risk is now all concentrated in just a few CCPs. We also don't know how the market will evolve after central clearing becomes mandatory. The markets have a habit of behaving with a certain ruthless efficiency when it comes to exploiting loopholes or opportunities created by regulation so, like the army that prepares for the last war, the new regulatory regime may not protect against emergent market risks and behaviours.
Bail-in is the power given to the regulator to declare a bank non viable and to impose losses on its creditor over a week end, and as a result auto-recapitalise the bank, which will be open for business and healthy the following Monday. You can see it as a flash, extra-judiciary chapter 11.
The regime is designed to impose losses on regular creditors (bond holders), rather than clients taking a credit exposure to the bank through derivatives or deposits, even if in a bankruptcy these would have the same ranking and should suffer the same losses. This should reduce a lot the disruption on the market of a bank going bad.
Banks have been required to hold minimum levels of bailinable wholesale debt to ensure regulators can do a large scale bail-in.
This a fundamental shift which banks, regulators, and large bank creditors are well aware of, but seems to be completely unkown to the general public. That's kind of as absurd as keeping the dooms day machine secret in Dr Strangelove. The ability to do bail-in is probably the main factor that reduced dramatically the risk of another 2008-style financial crisis.
They've taken the liberty of purchasing vast amounts of equities in the hopes of financing the Japanese retirement fund.
>While the Bank of Japan’s name is nowhere to be found in regulatory filings on major stock investors, the monetary authority’s exchange-traded fund purchases have made it a top 10 shareholder in about 90 percent of the Nikkei 225 Stock Average, according to estimates compiled by Bloomberg from public data. It’s now a major owner of more Japanese blue-chips than both BlackRock Inc., the world’s largest money manager, and Vanguard Group, which oversees more than $3 trillion http://www.bloomberg.com/news/articles/2016-04-24/the-tokyo-...
Demographics and central bank balance sheets being what they are, there is no way these purchases can continue indefinitely and this is certainly not indicative of a healthy market.
There are dozens of other problems which can cause the next one. Most regulations fight the last war, so the next one will be different.
> The simple fact is that anybody who thinks they know what is going to happen is dangerous, because they are messing with something that is very powerful that they don’t even remotely understand, or understand how it will change in response to meddling.
Or, more specifically, there are economic reasons why the network formed the way it should so we disturb it at our peril.
Huh?
First of all, this line of argument would seem to doom any attempt to regulate the economy whatsoever. One can always say: but the market is really complicated and it has developed the way it has for good reasons so its dangerous to mess with it.
He's right that this will always raise the possibility of Unintended Consequences. So it's a welcome point. But he's also discounting the fact that, given people's level of expertise, and the sometimes dismal status quo, it's hardly self evident that regulation is never the right move. Returning to this case: it's the obvious response that the network may have evolved the way it did for good economic reasons, but we have good reason to doubt that these reasons adequately include the safety of the financial market as a whole in the very long term? This, of course, is why we have regulators in the first place: because the aggregate interests of private actors in the market do not always correspond to our collective social interests.
> One major problem–which I’ve been on about for years, and which I am quoted about in the Nautilus piece–is that counterparty credit exposure is only one type of many connections in the financial network: liquidity is another source of interconnection.
> As a practical example, not only does mandatory clearing change the topology of a network, it also changes the tightness of the coupling through the imposition of rigid variation margining. Tighter coupling can change the probability of the failure of connections, and the circumstances under which these failures occur.
> Another problem is that models frequently leave out some participants. As another practical example, network models of derivatives markets include the major derivatives counterparties, and find that netting reduces the likelihood of a cascade of defaults within that network. But netting achieves this by redistributing the losses to other parties who are not explicitly modeled. As a result, the model is incomplete, and gives an incomplete understanding of the full effects of netting.
> Thus, any network model is inherently a very partial one, and is therefore likely to be a very poor guide to understanding the network in all its complexity.
FWIW, I think he's right to bring this up, as there are things that we already know of that the topology model doesn't cover.
Eg Black Monday in 1987 did not cause a recession. Neither did the flash crash. Even the Great Depression did not start because of the crash of 1929.
What's important is nominal GDP, and the central bank can target that to keep it on a stable growth path.
This should (IMO) be viewed as a decent summary:
http://noahpinionblog.blogspot.com/2012/07/excess-volatility...
I am not 100% sure I actually understand the post - I'd think that a "low pass filter" could be applied to Fed actions to address this, but it could easily be that I'm simply wrong about that. I have two competing biases - one, control theory, and two, a skepticism of how the EMH works out in real life.
Also, it's not clear to me exactly what would happen if the Fed simply adopted say, a 4% inflation target.
> Also, it's not clear to me exactly what would happen if the Fed simply adopted say, a 4% inflation target.
In practice, they seem to have an inflation ceiling, not an inflation target.
But it has not too much to do with nGDP targeting. (Or only insofar as nGDP targeting right now would suggest looser money, so going for more inflation would do the same thing.)
Also, are you talking about a 4% level target?
I generally read Sumner first. But I've caught that criticism of market monetarism from other sources. Now I wonder if it's valid.
I was talking about two things - nGDP and a 4% inflation target.
I wish we were talking about a 4% inflation target, although after watching the Yellen grilling a coupe of Mondays ago I wonder of we can even do that.
Stable nGDP growth seems pretty closely correlated with those two. Inflation less so.
The standard examples are: the oil price shocks of the 1970s gave us inflation, but nGDP crashed. The IT bubble and the housing bubble saw nGDP grow above trend, but not inflation.
Nominal GDP is relatively straightforward to define and measure; inflation comes with all kinds of thorny philosophical questions about quality improvements.
That being said, targeting the trajectory of the price level (=`level targeting') might still be a better idea than the current system, where the central banks see their inflation targets as ceilings instead, and don't make up this year for the targets they missed last year.
Scott Sumner argues that targeting aggregate nominal labour income might actually be slightly superior to targeting nGDP, but it's wages that are sticky. But the labour share of GDP is pretty stable over time in rich, diversified economies, so it doesn't make too much of a difference.
The best advertisement for nGDP targeting I've read was Scott Sumner's slogan that it's making the world safe for laissez faire capitalism: it's much harder to organise political opposition to creative destruction, if everyone knows that full employment is almost guaranteed, and there won't be any recessions. Let the banks fail, abolish tariffs, let the immigrants in, review restrictive licensing requirements, etc.
And your point about the oil shocks - and WIN buttons - also contributed to this.
I am currently reading Sumner's book about the Great Depression ("The Midas Curse"). Seems pretty interesting so far.
Now, I just need to find a country to migrate to that does nGDP targeting and finances itself via land value taxes alone.
Hint: I don't think Basic Income provides any such thing. I think it would worsen problems.
Though I think reducing the degrees of separation between the Haves and Have Nots can really help and the internet has enormous potential for facilitating that. However, we need to reduce prejudice and institute some best practices. Currently, the Haves do not want to hobknob with the Have Nots even online and even when being bludgeoned with an explicit expectation of getting help developing an income, not seeking charity. They just default to this assumption that poor people are beggars and charity cases and it goes bad places.
http://micheleincalifornia.blogspot.com/2016/07/money-is-not...
The point of basic income is similar to the point of minimum wage. Individual businesses cannot choose to raise their workers wages or they will be priced out of the market by their competitors. In contrast, if every company agrees to raise their minimum wage (or is forced to by the state), then the worker's share of wages improves. Of course this is not a permanent solution, since inflation reduces the utility of the minimum wage over time. Thus, basic income, like minimum wage, would need to be adjusted over time.
I also agree that money is not wealth, and I agree that the distinction is extremely important, particularly for understanding macroeconomics and ecological economics. Indeed, unsustainable divergence of money from wealth (through credit expansion of the money supply) is the prime cause of economic crises. However, under a properly controlled basic income system, such divergence could be minimized and even eliminated, thereby eliminating depressions.
So, while I agree that basic income cannot solve all our economic woes, I do believe it is a step in the right direction.
There is no upper limit to how expensive housing can get. Giving me $500 a month in Basic Income in no way guarantees that I still won't be able to afford housing.
The problem as I see it is that Bad Income is a lazy answer: oh, let's just cut poor people a check.
You skimmed what I wrote and dismissed it. You have said not one word about addressing the issue of the tremendous shortage of affordable housing, which has been steadily deepening.
So your reply in no way alleviates my fear that if we get Basic Income, the well off people involved in creating new housing stock will just have one more excuse to not care about solving the much thornier issue of affordable housing.
If there were affordable housing available, I could get off the street right now. Without it, I have no reason to believe Basic Income solves my problems and every reason to believe it deepens them.
Given the trends that go back decades, the affordable housing shortfall is the issue I would like to see tackled. And it is not happening.
Housing prices in California are high primarily for two reasons: a) California's population has quadrupled since 1950, and b) the US household debt to GDP ratio is ~twice what it was in 1950 [1], and higher household debt means higher housing prices [2].
By reducing the lower and middle classes' dependence on debt, basic income could actually reduce housing prices, or at least slow the rate of increase. Of course it's certainly feasible that basic income could take the form of housing subsidies. However, such an approach would add complexity to an otherwise very simple system, and the simplicity of the basic income system is one of it's biggest selling points.
All that being said, basic income does not address the issues of population growth, resource scarcity, and climate change. So, while I do think it would raise the average quality of life, I don't think it's sufficient to ensure our species success.
[1] http://www.financialsense.com/contributors/lance-roberts/con... [2] http://www.debunkingeconomics.com/us-flow-of-funds-update/
TLDR: New construction is about twice the size now as it was in the 1950s, with more amenities, while housing fewer people.
This goes a long way towards accounting for the higher housing prices alongside rising homelessness. You don't need fancy economic explanations. Houses that are nearly 2500 sq. ft. instead of 1200 and loaded up with vastly more amenities straight up cost more. Period.
And as long as new housing averages nearly 2500 sq. ft., no amount of basic income stabilizes the bottom third of the population because the housing they need simply does not exist. Period.
No amount of money gets you well if the cure you need simply does not exist. Ask anyone with an incurable medical condition.
Fist, I must say that upon reading your data points on average new housing size, my default mental response was "averages can be misleading". However, after looking up the data myself, it appears that the median house size has similarly nearly doubled in so much time. That is quite a remarkable figure, and I certainly agree that it has almost certinaly greatly raised the cost of living for those who wish to have their own space.
That being said, I strongly disagree that government-backed mortgages are the answer (not that you explicitly called for government-backed mortgages, but I think they are the consequence of your train of thought). Quite to the contrary, I would argue that the government-backed G.I. Bill mortgages, and similar programs in more recent years, are the root cause of the rising housing costs. As I described in my previous post, rising mortgage debt significantly increases housing prices by inflating housing asset prices. I think that, while not the be-all-end-all solution, basic income would be far more effective at countering homelessness than government-backed mortgages would be.
On a separate note, based on your descriptions of how basic income would affect the class dynamic, I think we have a different understanding of what a basic income would mean. Perhaps I should be more explicit in my wording. When I say basic income, I mean universal basic income, i.e. every citizen (and perhaps even non-citizens) would receive the same monthly dividend (or weekly, or biweekly). I fail to see how universal basic income would create a rift between the rich and the poor, since everyone would receive the same income. It is true that the poor would receive more utility from that income, but in my mind that disparity in utility is the main merit of the concept. That disparity in utility is what could transform our country from a land of Haves and Have Nots to a land of Haves and Have Mores.
On the topic of financial education, I complete agree with you. Universal basic income will not solve the problem of people with adequate income living month to month (or week to week). I don't think that's an argument against universal basic income, though. It's an argument for education reform. And by reform, I mean true reform, not charter school vouchers, and book purchase programs. We need to move from an education system designed to train factory workers to an education system designed to train service providers. The most important skills, skills that are, if anything, discouraged in our existing education system, are interpersonal communication and critical thinking. And financial education is right up there with those two.
But, I don't believe true universal income will ever happen. I think that if it gets implemented, it will basically be the new welfare.
Second, no, I have no desire to get government backed mortgages as the "solution" for affordable housing. I want an actual solution. I want more housing that is under 1200 square feet but not a trailer and not a Tumbleweed house on wheels.
In Japan, few homes have ovens. When I lived in Germany in my twenties, family homes all had refrigerators the size of what Americans put in dorms. These are first world countries with high standards of living, but they do not shackle their people with crazy "minimum" standards that cause the kinds of problems we have here in the US.
Tumbleweed houses -- one of the earliest companies in the tiny house movement -- were put on wheels as an end run around housing requirements that are supposed to prevent slums but just end up pushing people into trailers or out in the street. Our default expectation is a home so large as to fit a nuclear family, along with a giant fridge, a four burner stove and an oven.
When I got divorced and went from a family of four to a family of three, we began storing sodas in the fridge so the milk would not spoil. I assure you, a single person not only does not need a twenty cubic foot fridge, it is actively a burden for them.
Why are we more okay with the rising levels of homelessness than we are with small spaces with (say) a dorm sized fridge, hot plate and microwave? Your typical single person living alone has zero need of a four burner stove. Even with cooking for a family of four, I rarely used the fourth burner. I typically used two or three burners at one time. The result: The fourth burner became covered in grease and would smoke when I turned it on.
Why should I be saddled with a fourth burner I do not need and which will become a potential fire hazard if I do not clean it regularly, in spite of not using the damn thing? How does this prevent my home from being a slum? For me, a fourth burner is nothing but a headache.
And I cooked from scratch so much that the cashiers at the grocery store would say "I need you to talk to my wife. She never cooks. She's a microwave queen."
I was a full-time wife and mom at a time when that was out of fashion and I and my sons and husband all had special dietary needs. I cooked constantly. But I still only rarely used the fourth burner.
The housing standards we have are simply crazy. They are a burden for Americans, not something that protects us from abusive slum Lords. And they are increasingly pushing people out into the street.
I don't want a financial solution to this problem. I want a real solution that involves building more actual small spaces with basics instead of more and more McMansions while homelessness continues to rise.
And I think what this exchange tells me is that if I want to see real change here, I have my work cut out for me. Because even someone who will take the time to read my writing and treat me with genuine respect cannot immediately grasp that I am not talking about a fiscal solution. I am saying: We need to build millions of smaller homes with fewer amenities that ordinary people can afford without working three jobs and/or having roommates they don't really know. Because currently, the solution for unmarried individuals is rent a three bedroom place and get two roommates. And it simply should not be that way.
Thank you for engaging me in discussion.
Take care.
Then I would wait until people kind of forget the crisis and stop demanding banker's heads. 6-7 years? Then goto 1.
Disclaimer: not Mammon.
Since central banks can't unwillingly default on liabilities denominated in the currencies they control, this eliminates the second-order credit risk that was a major factor in the last crisis, when AIG and other institutions that issued credit insurance themselves became major sources of credit risk, first in the market for those instruments and then (due to the enormity of their losses) in the wider financial system. If a central bank must pay out on a CDS in the event of a crisis, the situation is little different from what happens anyway: central banks swell their balance sheets, effectively by printing money, to provide liquidity to the financial institutions. By acknowledging and formalizing this inevitable outcome through the issuance of credit derivatives (and by requiring major financial institutions to hold them), central banks can (a) make the financial sector pre-pay for its eventual bailing-out, (b) create a credible means by which to refuse further bailouts, and (c) facilitate a more effective and more informative market for credit instruments, which would help to prevent a crisis in the first place.
I suspect that this potential tactic has been ignored mostly because it would be politically infeasible to convince a financially unsophisticated public that CDSs—financial "weapons of mass destruction", in Warren Buffet's phrase—could be part of the solution. (But then it might be feasible: central banks are not accountable to the political process in the way that legislators are.)
At base, to tie this back into the submitted article, note that what I'm suggesting is little different from putting central banks into the role of central-clearing counterparties for credit risk, but with the added feature that these risks would be distilled into securities so that they can be priced and traded, with the concomitant benefits: the parties that receive credit insurance must pay for it, and markets for trading it aggregate otherwise diffuse information about the underlying credit risks.
Quite separately—at the same time I was musing about this stuff, I wrote a (rushed and rather mediocre) master's thesis that tried to expose the macro-level topology of the global financial system. The idea was to use the statistical correlations between major markets and assets classes to induce the structure of a metric space [0], revealing the "stochastic distance" between markets across time [1]. This ended up not being terribly useful, due to the well known fact that correlations between most markets move toward one in a crisis; the metric structure similarly collapsed in a crisis.
0. A metric induces a topology on the same underlying set.
1. In the course of my literature review, I found out that Rosario Mantegna had the core of the idea a decade before; see, Mantegna (2000), Introduction to Econophysics: Correlation and Complexity in Finance.
(My hazy mental model is that they are insurance; if that's even in the ballpark, the central bank has to price them correctly for them to be a tidy solution)
But, no, a central bank wouldn't really have to price its credit derivatives correctly. That's really the whole point: if it (and the market) gets the price wrong, the central bank can print money to fix the mess, which is (a) something that a private institution like AIG cannot do, adding a significant second-order risk that such institutions fail, and (b) when the market does get it wrong and a crisis results, the central bank ends up stepping in anyway, effectively issuing insurance post facto, only nobody has paid any premium for that insurance.
The trickier question is what notional value to issue. The obvious answer is that it would depend on each institution's contribution to systemic counterpart risk, but it might also make sense to leave the amounts to the discretion of central bankers: this would give them another lever of monetary policy to lean on, effectively giving them the ability to say, "No, don't worry about that bank failing. We've got you covered."
(And trickier still is how to ensure that the parties that are apt to suffer losses actually hold the insurance that would protect them. Particularly if it is expensive, they might not want to, and financial institutions have long created separate legal vehicles to obscure and lop off the risks they nevertheless indirectly retain.)
If we want to prevent the next crises, we need to address the underlying cause. I find Minsky's financial instability hypothesis quite poignant in this regard. I also highly recommend Steve Keen's work, which is heavily influenced by Minsky.
On Minsky—his thesis is that financial crises are inevitable. I've implied the same thing. Minsky proposed an antidote that relies on fiscal policy, which is politically determined and therefore unreliable (as we learned in the aftermath of the last crisis); I've proposed one that relies on what is essentially monetary policy wrapped in an automatic market mechanism.
First, the possibility of shorting gives people an incentive to dig up dirt early. Thus keeping management in line, and away from doing stuff that's too stupid.
Second (and less important), once someone shorts, they are forced to buy later. Thus, giving the stock a guaranteed buyer.
I suppose the real problem, then, is that stabilizing the stock system would require perfectly counterbalancing the two, which is infeasible in the real world. Stabilizing the system requires a financial damping friction.
For the same reason, I didn't see how traditional CDS's could realistically stabilize the system. I only saw how they could only change the oscillation period.
However, if the CDS's were issued by a central bank, which I now understand was the original author's suggestion, I do think I understand how they could play a fundamentally different role, i.e. the role of a financial damper. If the rate of issuance were properly controlled to be close to the critical damping condition, then I suppose the central bank could use such an instrument to stabilize the monetary system.
It might be an interesting research project to determine how the instruments would need to be priced to provide the desired damping.
However the Federal Reserve is thinly capitalized relative to the size of its balance sheet and as a result is extremely conservative in what it will hold on its balance sheet. Typically only holding bonds with Federal backing.
Any significant payout of CDS during a crisis would leave the Fed balance sheet insolvent and in need of recapitization.
That would require the Federal Reserve's member banks to recapitizie the bank via an equity buy-in which would further deplete member bank capital ratios which would already be under pressure in a crisis or a bail out by the Treasury.
Needless to say a bailout by the Treasury of the central bank would be politically untenable. People would be rightly concerned that the nations wealth had been committed to covering claims that in hindsight seem undervalued.
While there is plenty of central bank talk about helicopter money the central bank can't give money to anybody (only loan). Of course there is little difference between giving people money and providing a 1000 year loan at zero interest. The only politically tenable method of providing helicopter money is via monetizing Federal Debt as part of a spending package approved by Congress during a deep deflationary recession.
Anything else would likely bring out the torches and pitchforks.
It's true that the rules governing the Federal Reserve System and some other central banks date back to the days before fiat money, when cash represented a real claim on the central bank's gold or other assets. In that era, solvency of the central bank was a real concern. But today a central bank's capital balance has practically no meaning beyond the significance the central bank itself gives to it, and worrying about a central bank's solvency is rather silly.
That said, the credit instruments a central bank issues against the commercial banks in its jurisdiction should not represent uncompensated liabilities of the central bank. Aside from the premia collected by the central bank when it sells the CDSs at auction, the CDSs should also transfer equity in the underlying commercial bank to the central bank in the event of a credit event; in other words, the central bank would eventually be compensated by the shareholders of the commercial bank that has defaulted if the central bank must pay out on its credit instruments on that bank.
Again, the basic point is to formalize the informal mechanisms that already exist and are already begrudgingly employed ad hoc to bail out the financial system when a credit crisis occurs. The advantage of doing it through credit instruments issued by the central bank is that it dispels uncertainty about whether obligations will be paid, and about what will happen when they're not. The scheme eliminates uncertainty about who will be bailed out and who won't, and it creates markets that better reveal risks, while also managing to get the financial system to pay up front for at least some of its eventual bailing out.
It should also be possible to set the terms of the credit instruments to minimize their long-term effect on the central bank's balance sheet. Instruments with equity clawbacks and other provisions that would eventually push the losses onto the shareholders of commercial banks, making it unlikely that the central bank would suffer substantial losses in the long run. A credit crisis feeds on uncertainty: by issuing these instruments and paying out on them immediately, the central bank would prevent the financial system from seizing up, and thereby prevent a crisis of credit and liquidity from worsening and spilling over into the real economy.
Note also that the credit instruments I have in mind are intended to remove counterparty risk from the system. This means that they should give the holder a claim on the central bank in the event that a counterparty does not make a contractual payment; that is, they should be instruments on small credit events, not big ones like bankruptcy. Further, they should only compensate the holder for unpaid obligations to the holder. Thus the notional value of the instruments would not exceed the actual liabilities of commercial banks to their counterparties, and they would not be useful for speculation. Over the long term, the central bank would be on the hook only for actual unmet liabilities of the commercial banks it regulates, and if it's doing its job, those banks should be well enough capitalized that they are not in danger of becoming insolvent: the central bank would only suffer long-term losses if the commercial banks are truly insolvent, rather than if there is merely a temporal mismatch between their assets and liabilities.
Solvency of the Federal Reserve is required to control the money supply and manage inflation. When the Federal Reserve wants to put cash into the system it buys debt and creates cash. When it wants to take money out of the system it sells the debt and retires the cash it receives.
An insolvent central bank would cripple its ability to manage the money supply as it would never be able retire the cash for the defaulted asset because who would be willing (or required) to surrender their cash for the defaulted asset?
What you are arguing is essentially that the central bank should start operating as an insurance company instead of a central bank. Acting as the insurer of last resort so to speak. That function has traditionally (and rightly) been carried by the Federal Government and having the Central Bank take that role would require a dramatic change in its charter.
"In any party of six people either at least three of them are (pairwise) mutual strangers or at least three of them are (pairwise) mutual acquaintances"
https://en.wikipedia.org/wiki/Theorem_on_friends_and_strange...
I mean, despite the author's having worked on the street, he apparently never heard of correlation matrices.
No. It cannot.