Avinash D. Persaud gave a speech to the Committee of European Securities Regulators (posted at Willem Buiter’s blog) that argues that banks’ risk models and regulation based on market based pricing were bound to fail. That’s a very bold claim, yet Persaud appears to have the goods.
If any of you have worked with models, one of basic yet regularly-ignored rules is to understand and respect their assumptions, because they usually constitute a major limitation on their usefulness (the best remedy is to rely on multiple metrics and tools and apply good old fashioned human judgement, but many people prefer to default to the answer that pops out of a spreadsheeet).
Persaud tells us that an underlying assumption of “market sensitive risk models” is that the user is the only party taking that approach. Now instead of going to Zurich or the Caymans to coin money based on their findings, Harry Markovitz and George Dantzig instead made them public, which should have made them merely interesting. However, the practical implication was that they could be used successfully on a relatively small scale, but once they became common, their success became more and more erratic, as many quants and risk managers have learned to their dismay.
Persaud then launches another fundamental attack. He argues that the logic of regulation is circular. The purpose of regulation is to prevent market seize-ups (actually, I thought it was to prevent institutional collapse and damage to innocent and/or unsophisticated bystanders, but let’s go with his interpretation, since many have come to view well functioning markets as automagically producing those other results). But (and here Persaud is not as explicit as he might be) making market sensitive risk tools part of how financial institutions are regulated insures they will be widely, nay almost universally used, guaranteeing their failure. Eeek.
I would be curious to get the reaction of those skilled in the art. I’ve also included the reference to a paper by Persaud in 2000 which goes into his theory in more depth.
Sir Alan Greenspan, and others have questioned why risk models, which are at the centre of financial supervision, failed to avoid or mitigate today’s financial turmoil. There are two answers to this, one technical and the other philosophical. Neither is complex, but many regulators and central bankers chose to ignore them both.
The technical explanation is that market-sensitive risk models used by thousands of market participants work on the assumption that each user is the only person using them. This was not a bad approximation in 1952, when the intellectual underpinnings of these models were being developed at the Rand Corporation by Harry Markovitz and George Dantzig. This was a time of capital controls between countries, the segmentation of domestic financial markets and to get the historical frame right, it was the time of the Morris Minor with its top speed of 59mph.
In today’s flat world, market participants from Argentina to New Zealand have the same data on the risk, returns and correlation of financial instruments and use standard optimization models, which throw up the same portfolios to be favoured and those not to be. Market participants don’t stare helplessly at these results. They move into the favoured markets and out of the unfavoured. Enormous cross-border capital flows are unleashed. But under the weight of the herd, favoured instruments cannot remain undervalued, uncorrelated and low risk. They are transformed into the precise opposite.
When a market participant’s risk model detects a rise in risk in his portfolio, perhaps because of some random rise in volatility, and he tries to reduce his exposure, many others are trying to do the same thing at the same time with the same assets. A vicious cycle ensues of vertical price falls prompting further selling. Liquidity vanishes down a black hole. The degree to which this occurs is less to do with the precise financial instruments, but more with the depth of diversity of investor behaviour. Paradoxically, the observation of areas of safety in risk models, creates risks and the observation of risk, creates safety.
Quantum physicists will note a parallel with Heisenberg’s uncertainty principle.
Policy makers cannot claim to be surprised by all of this. The observation that market-sensitive risk models, increasingly integrated into financial supervision in a prescriptive manner, was going to send the herd off the cliff edge was made soon after the last round of crises*. Many policy officials in charge today, responded then that these warnings were too extreme to be considered realistic.
This brings us to the philosophical problem of the reliance of supervisors on bank risk models. The reason we regulate markets over and above normal corporate law is that from time to time markets fail and these failings have devastating consequences. If the purpose of regulation is to avoid market failures, we cannot use as the instruments of financial regulation, risk-models that rely on market prices, or any other instrument derived from market prices such as mark-to-market accounting. Market prices cannot save us from market failures. Yet, this is the thrust of modern financial regulation, which calls for more transparency on prices, more price-sensitive risk models and more price-sensitive prudential controls. These tools are like seat belts that stop working whenever you press hard on the accelerator.
My purpose is to explain why the reliance on risk models to protect us from crisis was always foolhardy. In terms of solutions, there is only space to observe that if we rely on market prices in our risk models and in value accounting, we must do so on the understanding that in rowdy times central banks will have to become buyers of last resort of distressed assets to avoid systemic collapse. This is the approach we have stumbled upon. Central bankers now consider mortgage-backed securities as collateral for their loans to banks. But the asymmetry of being a buyer of last resort without also being a seller of last resort during the unsustainable boom will only condemn us to cycles of instability.
The alternative is to try and avoid booms and crashes through regulatory and fiscal mechanisms designed to work against the incentives, fed through risk models, bonus payments and the like, for traders and investors to double up or more into something that the markets currently believe is a sure bet. This sounds fraught and policy makers are not as ambitious as they once were. We no longer walk on the moon. Of course, President Kennedy’s 1961 ambition to get to the moon within the decade was partly driven by a fear of the Soviets getting there first.
Regulatory ambition should be set now, while the fear of the current crisis is fresh and not when the crisis is over and the seat belts are working again.
*Sending the herd off the cliff edge: the disturbing interaction between herding and market-sensitive risk management models, A. Persaud, Jacques de Larosiere Prize Essay, Institute of International Finance, Washington, 2000.
Excuse me for my laziness, I’m cross-posting this comment from another blog.
I think it is probably incorrect to say that models “failed”. The models simply did not consider the scenario that we are living through.
There are *always* scenarios which will make a bank go bust. Consider a conservatively managed bank in the scenario where housing prices overnight go down 50% and people starting mailing in their keys. It would go bust. There’s no way around it, so long as a bank is the banking business.
Or consider the case of the Nasdaq bubble. At the height of the bubble if the market had gone down 80% in one day, then at least one IB would have gone bust. At least one, maybe more.
These are extreme scenarios, and they are not considered in the analyses that IBs run. At the height of the Nasdaq bubble suppose someone thought a scenario of -50% to be possible. Now, after the fact, we can say that it didn’t happen. Was the person wrong? Or was it a possibility that needed to be considered?
So banks can’t and won’t consider all possibilites. Some extreme scenarios will sink the bank; that’s not news. The hard part is to decide, beforehand, what is *plausible*.
For the proverbial “black swan” (or, for want of a better description, End of the World) events, the only thing that proves the plausibility of the event is if it happens. That makes it very difficult to know, beforehand, exactly what degree of scenario needs to be considered. At the height of the Nasdaq bubble what was the proper extreme scenario, for which a bank needed to be prepared? -25%? -40%? -50%? – 75%? -90%?
Now I don’t have any pity for IBs, but the current scenario is unprecedented. It’s a genuine “black swan” event. So post hoc it’s very easy to say, “Hey guys, you should have considered this.” But before the event, it’s very hard – no, it’s impossible – to calibrate.
So I’d say it’s not a question of tweaking the models. Post hoc there’s obviously ways to do that, but I imagine the IBs models are actually robust enough, and what was lacking was considering the present scenario as plausible, as an extreme event which needed to be considered and to be reserved against (and here the regulators, with their allegiance to VAR, were not helpful).
Persaud is fundamentally an apologist for things as they are. Herd behavior in financial markets isn’t exactly a keen insight, but insisting that central banks are responsible for cleaning up the inevitable bust is surely a comforting thought for people who are paid despite their failures of judgment. The problem isn’t that every mother’s son is using the same gaussian copula model, it’s that management is unwilling to ask questions when money is being made. Why pay attention to the model’s parameters if the money coming through the door seems to validate them? Models will never be better than management’s personal financial risk in relying on them. If you’re getting $160 million in any outcome, what do you care about the risk model?
What’s missing in the analysis is effect of duration mismatch – between management bonuses and the expected longevity of a financial institution. The risk from which the real world model was derived, from the viewpoint of a person compensated thus, is that the company fails to make lots of money now, not that the it might fail tomorrow.
Is there much else to this?
“Automagically”: that’s the best new adverb I’ve heard in many the day—and too true!
The issue with correlation can also be seen from connectivity processes in Game of Life simulations, which not only fit the context discussed here but are doubtless employed by some of the quant boys. Where connectivity is incomplete, you can have sustainable local optima—i.e. locally exploitable profit possibilities in this case—but where connectivity is high (every damn body’s got the same data and tools), than the system as a whole locks out differential valuations. If everyone gets into something, get out, is the only real lesson.
Risk assessment in the present circumstance has been totally faulty, but I would add another issue I don’t see in Persaud’s summary (or at least didn’t see, I’m up past my bedtime). If risk assessment is performed with the goal of avoiding risk, that outcome is more likely. If risk assessment is performed with the goal of taking on ‘manageable risk,’ then _that_ outcome is the more likely. The whole risk assessment game has been to find rationales for taking on as much as folks thought they could handle, not to avoid what they thought they couldn’t handle. This may sound small, but when every player at every level is looking to tack on just a litt-tt-ttle more risk, what you have is a system biased toward accumulating risk. There is a logical outcome to such a process; we live it today. It doesn’t matter, at a deeper level, whether the risk taker is someone with $50K on plastic and $0 in savings, a homeowner taking on an ARM above their cash flow with a plan to refi before the reset, or St. Alan the Mendacious believing he can bloom bubbles to grow the plutocracy ’cause he has the public till to clean up the mess and the mess will be small enough to be clean-upable. If the bias is to take on just a lit-tt-ttllllle more risk, one is guaranteed to get loaded with more risk than can be managed in an adverse circumstance. All that’s to say, risk management has been a fool’s game because the hidden bias has been to take it not to shake it. Except for Warren B., who treats risk like a disease and so will be the last man standing tall.
To a: No, the present situation is most assuredly _not_ unprecedented. Bubbles all look the same, and they are all readily recognizable on the way up. Look at SE Asia in the 90’s, Argentina, or any situation you want to pick. All it takes is to accept totally inflated values for assets as ‘normal’ and to use leverage in substantial excess of collateral. We have 400 years of modern capital doing the same damn thing, and they always do it the same damn way with a shiny new ribbon on the package. Just like we have.
It has been my view for some time that the only really stable public financial regulatory policy would be one _actively antagonistic to fast booms_ and deeply antipathetic to bubbles. Bubbles make big profits for big profiteers on the way up; this is why no one wants to prevent them. But they do massive damage. When we re-write the rules for public financial regulation in this country, preventing bubbles should be an explicit policy mandated to the Fed and public auditors. I’m not holding my breath on this one, though. Jus’ too many folks want to get rich rather than live well. It isn’t zero sum, this ‘get rich’ thing, but it tends to sum to zero repeatedly.
An analogy for taking risk which I left off the previous post is this:
Two bottles of vodka can kill you if you drink them all at once. If you try, you will vomit, and so save your life. If you are a hardcore drinker and try to chug them, you’ll probably still upchuck and live to try again. If you’re hardcore and take it slow, though—knock a shot down, talk a bit, take another, repeat—you can get a dose plenty enough to kill you while you are still upright and holding (mostly) sensible conversation. Even if you ‘know your limit’ it can change on a give day; you’re more dehydrated to start, going a little too fast, etc., etc. It doesn’t have to be a big belt that embalms you.
Now, our systems was _supposed_ to be that the barkeep would cut off the hardcore drunk before the dose got toxic. But, in our situation the pourer was a pusher. If you learn to ‘manage risk’ like you can learn to ‘hold your liquor,’ you have a great chance of getting enough in you to be a dead man talking. Shake hands with anyone on Wall Street today looking for more capital and asset price support and you’ll know just what I mean.
“Paradoxically, the observation of areas of safety in risk models, creates risks and the observation of risk, creates safety.”
That sounds like Minsky in a nutshell.
“Automagically”: that’s the best new adverb I’ve heard in many the day—and too true!
You obviously don’t hang with computer guys. “Automagically” dates back to the nineties or before. Usually refers to some hacked up initialization code.
CHECK it out!!
Two U.S. senators, citing a plunge in regulatory fines against companies and individuals, yesterday asked government watchdogs to examine whether the federal Securities and Exchange Commission isn’t being aggressive in fighting fraud.
Sen. Jack Reed said in a speech that he and Senate Banking Committee Chairman Christopher Dodd, D-Conn., asked the Government Accountability Office to look at the SEC’s enforcement division. Reed sought the review after sanctions fell about 51 percent to $1.6 billion in fiscal 2007.
The drop “raises questions about whether changes have taken place in enforcement philosophy or scope of activity,” Reed said in a statement. The GAO will examine whether the enforcement division “has sufficient staff and funds to perform its mission and whether there have been fundamental changes in operation to the way they handle cases.”
This is the second time in 18 months that a member of Congress has called for a review of the SEC’s vigilance. Since taking over in August 2005, SEC chairman Christopher Cox has given commissioners more power over enforcement penalties as the agency considers whether fining companies also hurts shareholders.
Isn’t this just another version of Goodhart’s law?
Danielsson even has a corollary to that as “A risk model breaks down when used for regulatory purposes”
See this paper
The Emperor has no Clothes: Limits to Risk Modelling