You give me a capital requirement, I’ll give you a derivative to skirt it.
I’ve enjoyed reading Anat Admati and Martin Hellwig’s recent book, The Bankers’ New Clothes, which explains a ton of things extremely well, including:
- Differentiating between what’s “good for banks” (i.e. bankers) versus what’s good for the public, and how, through unnecessary complexity and shittons of lobbying money, the “good for bankers” case is made much more often and much more vehemently,
- that, when there’s a guaranteed backstop for a loan, the person taking out the loan has incentive to take on more risk, and
- that there are two different definitions of “big returns” depending on the context: one means big in absolute value (where -30% is bigger than -10%), the other mean big as in more positive (where -10% is bigger than -30%). Believe it or not, this ambiguity could be (at least metaphorically) taken as a cause of confusion when bankers talk to the public, in the following sense. Namely, when the expected return on an investment is, say, 3%, it makes sense for bankers to lever up their bets so they get “bigger returns” in the first sense, especially since there’s essentially no down side for them (a -30% return doesn’t affect them personally, a 30% return means a huge bonus). From the perspective of the public, they’d like to see the banks go for the “bigger return” in the second sense, so avoid the -30% scenario altogether, via restrained risk-taking.
Admati and Hellwig’s suggestion is to raise capital requirements to much higher levels than we currently have.
Here’s the thing though, and it’s really a question for you readers. How do derivatives show up on the balance sheet exactly, and what prevents me from building a derivative that avoids adding to my capital requirement but which adds risk to my portfolio?
I’ve been getting a lot of different information from people about whether this is possible, or will be possible once Basel III is implemented, but I haven’t reached anyone yet who is actually expert enough to make a definitive claim one way or the other.
It’s one thing if you’re talking about government interest rate swaps, but how do CDS’s, for example, get treated in terms of capital requirements? Is there an implicit probability of default used for accounting purposes? In that case, since such instruments are famously incredibly fat-tailed (i.e. the probability of default looks miniscule until it doesn’t), wouldn’t that encourage everyone to invest extremely heavily in instruments that don’t move their capital ratios much but take on outrageous risks? The devil’s in the detail here.
mathbabe 
I can’t provide details, but it seems obvious that it *is* possible to set up adequate capital requirements for a CDS. A CDS is financially equivalent to an insurance contract (not unlike catastrophe insurance), and the insurance industry has managed to write lots of insurance policies without many systemic crises. If you simply imposed the same kind of accounting and capital requirements as for insurance companies, then the systemic risk drops dramatically. The problem is that all the excitement and profit margins of Wall Street would also go away, and investment banks become quiet and sleepy places (like they once were and like insurance companies). This seems unlikely to happen.
LikeLike
No doubt it’s possible for any single derivative. I guess my point is that it’s a feedback loop, and new derivatives will show up that haven’t been invented yet.
LikeLike
Insurance companies are also always trying to invent new types of insurance, but this is tightly regulated. In principle (if not in practice), the same thing could be done with derivatives. If you focus on simply defined policies that limit the risk and forget about protecting profit margins, it all gets very easy. What forces things like Dodd-Frank to become thousands of pages long is an effort to “reduce risk” without reducing profit.
LikeLike
I do not claim to be an expert but I am quite confident that it is possible to game Basel III rules. In any case, no one can sensibly assert that it isn’t. The best they could say is, “As far as I know, no one has figured out how, yet.”
Recognizing that no system will be perfect, doesn’t undermine the case for substantially higher capital requirements. We shouldn’t let the perfect be the enemy of the good.
If anything, it makes the case stronger. If you are concerned about gaming of the regulations, don’t you want a large cushion?
The clearing and disclosure rules will help also. If (when) they find ways to game Basel III (or IV, or V) it is better if it is visible.
In any case, we should also work to have comprehensive regulations. And, of course, regulators who are alert and not compromised would be extremely helpful too
LikeLike
Agreed on all points. I’m not saying don’t have higher capital requirements – by all means do. I just want to know how well I can sleep at night if that’s all we do.
LikeLike
I don’t think insurance regulators are any more alert than financial ones. If anything, they’re a lot sleepier. I believe the idea that Wall Street regulators need to be smart and savvy only promotes the concept that financial regulations have to be sophisticated and complex. This in turn simply leads to more loopholes to exploit. I believe it is possible to have short and simple regulations that would achieve the right goals (mainly, little or no systemic risk) much more effectively than “sophisticated” regulations and regulators.
LikeLike
Great points!
LikeLike
@Deanne:
I am all for simplicity, too. That’s part of the reason I like capital rather than the Volcker Rule.
For something I am writing, I would be delighted to hear other suggestions for “short, simple” rules.
Having said that, I do think alert regulators are needed. The market will inevitably change and people will think of new ways to game it, there needs to be a response.
LikeLike
The Devil in the Details is the deliberate structured complexity enabled by the “deregulation” that made activity that used to be clearly criminal (and beget shame, financial banishment and imprisonment) is now standard operating procedure, with mere cost-of-business-slaps on the wrist for outright fraud. Such complexity can only be cured by again recognizing that financial fraud is not at all self-regulated by market forces, it is driven to extremes by the need to compete with those profiting most from getting away with the latest obscurantist crooked gimmick. The simplicity of criminalization is neither quaint nor out-of-date — it is VITAL to the survival of anything claiming to approach a free market.
LikeLike
Isn’t the cleanest and most simple “regulation” to shrink the banks to where they would be allowed to fail and all the parties know it? Doesn’t that cut the Gordian Knot? There would be little point to gaming because the games would put the bank in real jeopardy. Ok, I might be persuaded to add claw-back provisions for the compensation of senior executives. That might be the only regulation necessary. If bond and equity holders where actually at risk and knew it, the price feedback loop would be more real and effective.
Having said that, I do not for a minute believe that our intellectually and financially captured political system will defy its master — though I am always happy to be proven wrong. In the meantime, that is, as some souls continue to advocate and fight for smarter regulation to reduce the risk to “the people,” I believe that a campaign to divest (banking and investing) from the Big 6 is the peoples’ end-around. If our elected representatives won’t protect us from “too big to fail” (and it appears that they will not), then the people themselves have to do what their castrated representatives will not.
And that means educating people that the TBTF policy is like the sword of Damocles hanging over our heads. Our representatives are holding us and the next generation hostage by allowing a policy that distorts market mechanisms whcih are designed to rein in risk. This policy has not been put forth to the people to vote upon and is a major exception to the basic rules of capitalism and markets. It states that if these institutions THAT ARE ALLOWED TO GROW TO BE TOO BIG TO FAIL, in fact fail, then our representatives promise to steal from us and our children whatever is necessary to prop us flawed and poorly run businesses. In the meantime, these institutions are incented to grow ever bigger and ever more connected in order to both benefit from the economic gains of such status and to preserve and increase their hostage-taking status.
I do believe that there are simple regulations that would work. I do not believe they will pass. The shitton of money will see to that.
Banking and investing with the Big 6 is like remaining in a burning building when all you have to do is walk out. People need to know that they are being set up for another round of theft. Fool me once, shame on you……
LikeLike
Yes, getting people to take their money (and credit card accounts) away from TBTF banks would be good and, as you point out, something we can do now without any help from the government.
Shrinking the banks would be an improvement but it isn’t a panacea either.
Size is less the problem than connectedness.
The money-market-fund industry was bailed out even though none of them were big.
LikeLike
Great points. Thanks for the “credit card” add. I forget that sometimes when talking to people. Also, yes the interconnectedness. Not sure what one can do about that but if they were smaller there must become a point where they would be small enough to fail despite interconnectedness. IF banks knew they not only they would be allowed to fail but there counter parties would be, as well, might they decide to be a little more unconnected or, better yet, serve in the role of bond vigilantes?
Money markets should NEVER EVER have been bailed out. What a terrible feedback loop that was.
LikeLike
For those who may not have time to read the whole book, Admati was interviewed by Russ Roberts of EconTalk recently (just an hour) http://www.econtalk.org/archives/2013/04/admati_on_bank.html. And here’s the link to the great interview with Altbanking’s Cathy O’Neil if you haven’t listened to it: http://www.econtalk.org/archives/_featuring/cathy_oneil/ And Charles Calomiris: http://www.econtalk.org/archives/_featuring/charles_calomir/
LikeLike
Have you read something like “Moderninsing Money,” by Jackson and Dyson? It’s about banking reform, specifically in the UK, but the US system seems fairly similar. I am in new york, and would lend you my copy if you are interested. One sentence summary: remove money creation ability from the commercial banks and give it to the central bank.
LikeLike
Just saw this on Occupy’s FB page: My point. http://on.fb.me/12dZAnp
LikeLike
First off, I’m no banking expert. There are undoubtedly plenty who can explain technicalities well. I am therefore happy to stand corrected on operational and technical detail by those with more experience.
But having worked at AIG (on the regulated insurance side) when it was saved, i have a different perspective on the key differences between regulated insurance & “unregulated” ISDA “insurance”. The former fully funds risk – the insurers balance sheet is essentially statutorily protected for policyholders when an insurer fails. ISDA “insurance” (wrt credit risk in particular, mainly purchased Over The Counter but moving slowly to Central Clearing Counterparties), currently offers fewer statutory guarantees for corporate end-users.
First things first, derivatives are classed as credit risk mitigants – similar to guarantees.
http://www.bis.org/bcbs/qis/qis3qa_e.htm
In theory, Banks that purchase protection/CRM’s to reduce their credit exposure must ensure they satisfy the operational requirements defined by the BIS in order to be eligible for recognition from an economic and accounting perspective; CRM’s must be direct, explicit – linked to definite exposures, clearly defined and incontrovertible, irrevocable, legally enforceable in relevant jurisdictions, lack conditionality outside the control of the purchaser, etc (see page 33 http://www.bis.org/publ/bcbsca04.pdf -superceded document but definitions still OK).
I said “in theory” because the cases of Bankia and Greek bank default to name but two, have shown how derivatives may fail to legally perform as the purchaser expected them to. This can not only be financially ruinous to a purchaser but also suggests banks may be understating their RWA’s. This is because they are all too often assuming they will economically and legally perform using the pre-2008 standards; a pretty big assumption about the historical diligence of our legal professionals. With insurance, if you don’t smoke you pay lower premiums than a smoker; derivatives are less sensitive to human choices and behaviour and are more linked to financial decisions often within the control of financial markets (exogenous risks to the Real Economy company relying on the financial markets use of derivatives). As you suggest, this may discourage prudent behaviour.
Banks selling protection must record a liability, so the entity purchasing the protection must follow the process above.
Once satisfied that derivatives are operationally compliant they can be financially recognised on the Balance Sheet. The Exposure at Default and/or Loss Given Default percentages that relate to the anticipated future cash flows for each asset (defined in each derivative) whose default risk is being mitigated can be reduced to take economic account of the “guarantee” expected to perform.
Banks therefore have a Pre-CRM Risk Weighted Asset Exposure Amount and a post-CRM RWA EA. Supervisors should be looking closely at both. Clearly there has been some gaming going on which resulted in this from the BIS in 2011 regarding abuse of CRM’s -http://www.bis.org/publ/bcbs_nl16.htm . Outside the US, how banks mitigate its risks are published in the Pillar 3 Disclosure released annually with its full year financials (check out Barclays Pillar 3 Disclosure 2012).
In terms of the accounting treatment there’s a great deal of ongoing debate about the merits of the FASB treatment (where Netting agreements between counterparties enable banks to present NET rather than GROSS exposures) versus IFRS where Gross exposures are disclosed). Needless to say,
I could go on about hedge accounting and the London Whale use of derivatives to “hedge’ balance sheets but should probably leave it there. Enough for the time being.
LikeLike
I would have to heavily disagree with this statement “the insurance industry has managed to write lots of insurance policies without many systemic crises”, just look at what happened to AIG! Lloyds was actually a big Ponzi-scheme, British insurances used to sell annuities with guaranteed interest rates of more than 6% and there are plenty more examples, especially if you look at structured products and reinsurers. Furthermore the regulation momentarily in place (at least in Europe this is Solvency I) is woefully inadequate in dealing with modern insurance products (variable annuities, unit-linked business etc) . Ironically the new insurance regulation (Solvency II) in Europe is heavily based on the ideas from the Basel framework.
LikeLike
Cathy,
The short answer is ‘yes’. You can game absolutely anything, and the more precise the rules are, the easier it is to game them.
With specific reference to Basel, it looks like there will be a host of unintended consequences – and therefore opportunities to practice regulatory arbitrage. A dollar of CDS will count as more than 10x the asset weight of a dollar of a swap on the equity of the same company – despite the fact that in most people’s minds, equity is riskier than debt.
You can point to similarly lunatic ideas about how collateral will be counted, and that’s before you get to netting your exposures (by instrument, counterparty, or whatever). In fact, the treatment of derivatives in general is less than completely intuitive.
It’s easy to make fun of Basel, just like it’s fairly easy to make fun of most complicated things that have been put together by well-meaning people. Basel is in a bit of a difficult spot, however: non-bank agents (corporates, people, whomever) wants to own things that are safe and put their risk in their borrowing. Someone else – banks and their shadows – has to take the other side of that trade. By definition, banking is a risky business – and de facto, a complicated one – and regulating it efficiently, so that the tension between what bank shareholders can reasonably expect and what society as a whole might expect at any given time is resolved, is a tricky undertaking.
LikeLike
To answer the question, both:
Commercial bank net interest margins, default risk, interest-rate risk, and off-balance sheet banking
L Angbazo – Journal of Banking & Finance, 1997 – Elsevier
… The relationship between off-balance sheet banking and on-balance sheet portfolio risk is
examined by testing the hypothesis that off-balance sheet exposure has no effect on default risk, interest-rate risk, capital risk, or liquidity risk. Derivatives may affect risk in the commercial …
And more here (stochastics):
http://mailer.fsu.edu/~jclark/research/jrev0201.doc
LikeLike
On the question “what prevents me from building a derivative that avoids adding to my capital requirement but which adds risk to my portfolio?”
According to Anthony Saunders and Viral Acharya of NYU risk-weighting is still based on portfolio VaR (among other things, but VaR seems to be the principal measure) — both recent daily VaR as in Basel II, and stress VaR (Basel III) “based on a 12 month Period (250 days) of a period of continuous stress (e.g., 2007, Russian crisis etc…)” (http://pages.stern.nyu.edu/~sternfin/vacharya/public_html/Dodd-Frank-Basel-and-India-by-Viral-Acharya-2.pdf, p 25).
You can drive trucks through this if you have a mind to (and where there’s a will there’s a way…). All you need to do is construct trades so they survive the stress VaR test for some period of continuous stress. A little data mining and data-fitting should suffice.
If you pursue this I would suggest contacting Saunders or Acharya. They appear to know what they’re talking about.
LikeLike
I fear that we’re trying to solve the wrong problem here. The wrong problem is how, given light-touch regulation, we can stop banks from taking absurd risks via derivatives or otherwise. It’s wrong because bankers are strongly motivated to beat the regulators and are more numerous and probably smarter.
Capital requirements help but won’t be enough. One (partial) answer is to separate ordinary from casino banking. A Tobin tax can then take the heat out of casino banking – in fact imo it should be used to shrink that sector dramatically. Ordinary banking should then be a tightly regulated sector. In the short-term (5 years?) all innovation and product change should require prior agreement from regulators who should approve only innovations that (1) they understand, (2) they deem safe and (3) benefit customers and/or society.
In parallel ordinary banks should be required to conduct major culture change programmes to embed a customer service ethos. Once the regulator is satisfied that the change is embedded the ordinary banks can be given more flexibility and allowed to make some changes without prior agreement from the regulators.
LikeLike
A Tobin tax yes, if you can get the details right. And the rest is great too with the sunset provisions you seem to imply since the pendulum has swung too far toward industry self-regulation. In the longer run one should be as skeptical of government as of banks.
Another approach is to think only of how to protect the public against the direct danger to it, which would be of a complete breakdown in the financial supply chain (you can’t get your money out), or a high enough probability of it to cause panic. A solution would be for government to be credibly prepared to backstop that supply chain if needed – a FEMA for the financial supply chain.
LikeLike
I am still inclined to believe that the best ideas in the world don’t have a chance of passing. And if something does pass, it’ll only be because the bankers know in advance that there is a way around it. I guess that’s textbook cynicism but they have the money, the influence, the incentives, and inventiveness.
The people only have this vague fear or sense that something is rotten in the state of Denmark and are unlikely to rise up unless somehow agitated.
We is need is a Frank Luntz-like Denis ex machine to ride down with the perfect framing. Rage itself won’t work, in my opinion, UNTIL the next crisis. And even that didn’t work this time.
That’s not to give up or to stop pushing on the legislative front. To roll over would be even worse.
At altBanking, someone (Polar?) harnessing suggested networks. One group he mentioned was Churches. I think there is an ally in the new Pope. His first encyclical (the one released today was a finishing of Benedict’s last one) is reported to be about “the poor.” Care for the poor, preferential option for the poor, the dignity of the poor. I doubt it’ll be long but we may want to be prepare in advance to work with the American Bishops. Pope Francis has already changed the focus and tone of the Catholic Church.
There is no question that poor are being taken advantage of by financial institutions.
I am also still bewildered about why community banks and credit unions aren’t making hay with this and crying foul.
Wasn’t it the smaller banks that banded together to push FDIC insurance through so they could compete effectively with the larger banks that seemed more safe and sound.
What do the larger banks have over them in terms of power. It is so anti-capitalist for them not to be screaming about how unfair the advantages are to the big guys. Does anyone know?
LikeLike
See, silly me, I thought this was an accounting question: you can always circumvent how anything counts for capital regulations by keeping it off-book. I’m also not sure what to do with the concept that, even if you count certain delineated “assets” as required by Basel, when other assets on the books are being marked to fantasy based on another set of rules, the net effect on your reserves is whatever you want it to be. So long as there is any class of risks/assets not covered by a uniform set of rules, there is always interstitial space between sets or rules and between silos in a balance sheet to put an entire air force carrier into. And we all know how easy it is to create a new kind of paper – the whole idea of derivatives is that they exist like a fractal shadow of some real asset, spiraling off into infinite risk. It might look like insurance, but it quacks like doubling down at the casino.
LikeLike
Here is someone else’s educated response to Admati.
http://www.bankstocks.com/ArticleViewer.aspx?ArticleID=6660&ArticleTypeID=2
“At the depths of a financial panic, it’s not inadequate capital that usually dooms banks, it’s inadequate liquidity.”
On a separate point, using VAR to analyze portfolio risk is as good as using Black-Scholes in an 18-sigma event. Boggles the mind that intelligent people would use either one.
Happy Independence Day. God Bless America.
(sent from sunny Haifa, Israel)
LikeLike
You have a point there — sure makes you wonder.
LikeLike
Illiquidity typically arises because short-term lending or deposits are being pulled because of questions of solvency. And, it is easier, and less costly, for a central bank to help if there are adequate assets.
To suggest that 2008 would have been just as scary and damaging if the banks came into it with 10 times the capital is silly.
Panacea, no. Big improvement, yes.
LikeLike
Illiquidity in the case of Lehman arose because of a number of factors, but having 10 times capital would not have made a difference. Hint: it helps to have your top level employees rotate to jobs in government who then decide who to save and who to let flounder. Dick Fuld could not have been any lonelier.
LikeLike
From today’s Kwak and Johnson “Baselinescenario” blog. In it, they suggest that one of the real probalems is that “capital cannot be measured.” They point out that Lehman had 11% Tier 1 capital two weeks before it went bankrupt. If bank capital indeed cannot be measured and can be circumvented, there seems to be a problem of a “limiting factor.” As Josh says, higher capital would be a big improvement, but it’s not clear that higher capital “requirements” would lead to higher capital.
And even if our representatives or The Fed would demand sufficient (Admati/Hellwig) capital requirements — which they will not, as we see — it just makes failures “less likley.” I honestly don’t want a dime of my money stolen in future taxes because our representatives give banks an exception to the rules of capitalism. It’s one heaping bundle of moral hazard if you ask me. Until the individuals who run the banks not only know that their enterprises will fail if they are not prudent (like every other business) and that they will pay personally for that failure, they will find ways to continue their “heads I win, tails you lose” strategy. It’s actually the rationally immoral thing to do.
LikeLike
I like the Baseline Scenario guys a lot. But I don’t think they mean that capital doesn’t matter.
http://www.bloomberg.com/news/2013-03-03/why-higher-bank-equity-is-in-the-public-interest.html
LikeLike
No. No. I didn’t mean that. I agree. They would definitely agree with higher capital requirements for exactly the reasons you have mentioned. And Admat/Hellwig. I just meant that they recognize that it can be gotten around in a variety of ways, including that it is so hard to measure.
LikeLike
I posted again – good discussion!
Sent from my iPhone
LikeLike
I suspect the only ways to make people more cautious about the risks they may be putting on is to make sure they have downside skin in the game. If your govt-supported bank goes bankrupt, any C-level executives lose most of their assets and potentially have some jailtime.
LikeLike
Yes. Skin. Always. If not, then the risks are always worth taking.
LikeLike
Derivatives are capitalized according to ‘current credit exposure’ and ‘potential future exposure’ measures. Derivatives exposure measures and netting permitted are one of the most complicated areas of capital rules and an area where banks focus a lot, sometimes more than on the baseline capital requirement itself. Some of reform advocates (including Americans for Financial Reform) have called for using IFRS gross derivatives measures because it doesn’t allow games with netting. The PFE stuff is another case; the standardized lookup tables for PFE seem inadequate but it is hard to know what to suggest instead. E.g. the PFE that has to be capitalized on credit derivatives is just a fraction of the total amount potentially at risk.
LikeLike
Marcus,
Awesome! I need to sit down and work through some examples with you. That would really help me a lot!
Thanks,
Cathy
LikeLike
Would you and Marcus please write up your conclusions (even if they’re not conclusive) in a follow-up post? I revisited the comments to this post after reading Admati and Hellwig’s response to their critics:
Click to access parade-continues-June-3.pdf
In it, they don’t respond to the criticism that capital requirements can be gotten around with derivatives or (as the Dallas Fed argued) by putting risk on the balance sheets of subsidiaries. Any clarification about the issue would be appreciated.
LikeLike
Let’s take a simple case: I’m long a Sep 250 Call, short a Sep 250 Put, short a Sep future, all on the same underlying, all expiring at the same time.. My net exposure is 0. Should my capital requirement be calculated by taking margin on each instrument separately? Clearly, in this case netting makes sense. But what if I replace the 250 Call with a 255? Or perhaps a 300 or a 200? (As for future exposure, VAR is a joke.)
LikeLike