Epicurean Dice

The Epicurean Dealmaker has a post about risk and uncertainty. He makes some good points, and I want to expand on one of them, that is the respect we should have for the random nature of the markets.

Think about it like this. Mostly in finance we assume that we have the equivalent of a standard dice. That is, while we assume we don't know what number will come up next, we think that we know the distribution of numbers perfectly. In fact the real situation is much more akin to throwing a dice where we have imperfect knowledge of what numbers are on the faces. They might be 1 to 6; but they also might be 1 to 5 with the 1 repeated; or 2 to 7; or something else entirely. Worse, the numbers are changed by the malevolent hand of chance on a regular basis. Not so often that we know nothing about the distribution, but often enough that we cannot be sure that the current market will be like the past.

Thus our risk estimates are potentially wrong for at least two reasons. We might have been wrong about the past distribution. And even if we got that right, it might be different in the future. In other words, you can't manage risk effectively by assuming you know the distribution - to be effective, you really must assume that you don't. Thus you don't just want your risk to be low enough based on one model: you want it to be low enough based on all (or at least all likely) models.

California Foreclosures

Prompted by an article in the WSJ on foreclosures, I did a little research. The basic issue is to what extent banks are delaying selling foreclosed inventory, or delaying the foreclosure process, either because they do not want to realise the loss, they do not want to increase their volume of REO (real estate owned) or they do not physically have the capacity to process all the foreclosures they have. So, how long can a bank delay an auction once the property has been foreclosed? In CA at least, the answer is a year. The relevant portions of the CA state code are here, if you have tolerance for US law, and the key paragraph reads:

There may be a postponement or postponements of the sale proceedings, including a postponement upon instruction by the beneficiary to the trustee that the sale proceedings be postponed, at any time prior to the completion of the sale for any period of time not to exceed a total of 365 days from the date set forth in the notice of sale. The trustee shall postpone the sale in accordance with any of the following:

1. Upon the order of any court of competent jurisdiction.
   
2. If stayed by operation of law.
   
3. By mutual agreement, whether oral or in writing, of any
   trustor and any beneficiary or any mortgagor and any mortgagee.
   
4. At the discretion of the trustee.

So we are in a situation where property is dripping out of the bottom of the bucket through foreclosure sales, property is entering the bucket through delinquencies turning into foreclosures, and property cannot stay in the bucket for more than a year, ordinarily. This is clearly going to lengthen the duration of the real estate downturn, at least until the bucket starts to empty. Clearly it is in banks' economic interest to get this done, but I wonder whether they have the capacity to sell at a much faster rate than they are doing now, or whether the market in parts of California, Nevada, Florida and Illinois will support that volume of sales.

Update. Foreclosure volumes are rising fast in CA: 1,300 a day are now being executed according to the LA Times, or more than three times the rate of a year ago. I'd really like to see a detailed industry wide analysis of the levels and trends in delinquencies of various ages, foreclosures, and REO to get a sense of how full the bucket is. My sense is that the banks are effectively long rather more property than they would have us believe but pinning this down is difficult.

Another update. Further background is here (an LA Times update) and here (a discussion by Sacramento Real Estate Statistics of a Deutsche research report on Shadow Inventory).

My book has just been published

Finally Understanding Risk: The Theory and Practice of Financial Risk Management has appeared. Here's a link for amazon. The high level contents follow:

Part One: Risk Management and the Behaviour of Products
Chapter 1 Markets, Risks, and Risk Management in Context
1.1 Financial Markets Overview
1.2 Trading and Market Behaviour
1.3 Basic Ideas in Risk Management
1.4 Culture and Organisation
1.5 Some External Constraints

Chapter 2 Derivatives and Quantitative Market Risk Management
2.1 Returns, Options, Sensitivities
2.2 Portfolios and Risk Aggregation
2.3 Understanding the Behaviour of Derivatives
2.4 Interest Rate Derivatives and Yield Curve Models
2.5 Single Name Credit Derivatives
2.6 Valuation, Hedging and Model Risk

Part Two: Economic and Regulatory Capital Models
Chapter 3 Capital: Motivation and Provision
3.1 Motivations for Capital
3.2 Capital Instrument Features
3.3 Regulatory Capital Provision

Chapter 4 Market Risk Capital Models
4.1 General Market Risk Capital Models
4.2 Some Limitations to and Extensions of Value At Risk Models
4.3 Risk Systems and Risk Data

Chapter 5 Credit Risk and Credit Risk Capital Models
5.1 The Banking Book: Introducing the Products and the Risks
5.2 Credit Risk for Small Numbers of Obligators
5.3 An Introduction to Tranching and Portfolio Credit Derivatives
5.4 Credit Portfolio Risk Management
5.5 Political and Country Risk

Chapter 6 Operational Risk and Further Topics in Capital Estimation
6.1 An Introduction to Operational Risk
6.2 The Tails and Operational Risk Modelling
6.3 Allocating Capital and Other Risks

Chapter 7 Bank Regulation and Capital Requirements
7.1 Regulatory Capital and the Basel Accords
7.2 Basel II: Beyond the capital rules

Part Three: Treasury and Liquidity Risks
Chapter 8 The Treasury and Asset/Liability Management
8.1 An Introduction to Asset/Liability Management
8.2 Banking Book Income and Funding the Bank
8.3 ALM in Practice
8.4 Trading Book ALM

Chapter 9 Liquidity Risk Management
9.1 The Liquidity of Securities and Deposits
9.2 Liquidity Management
9.3 Contingent Liquidity and Contingent Funding
9.4 Stresses of Liquidity

Part Four: Some Trading Businesses and their Challenges
Chapter 10 An Introduction to Structured Finance
10.1 Contractual Relations
10.2 Asset Backed Securities
10.3 Securitisation Structures and Technology

Chapter 11 Novel Asset Classes, Basket Products, and Cross Asset Trading
11.1 Inflation-linked Products
11.2 Equity Basket Products
11.3 Convertible Bonds
11.4 Equity/Credit Trading
11.5 New Products

When is safety a good idea?

Faced with that title, the experienced reader will immediately say 'it depends what you mean by safety'. Let's start with a seemingly innocuous proposition: 'if you can make something safer at no or little cost, you should'. One potential counterexample here is bicycle helmets. While the evidence is by no means definitive, it does suggest that making people wear helmets makes cycling less safe. What seems to happen is that people feel safer when wearing a helmet and hence take more risks. They are not that much safer, so the extra risks they take introduce more risk than the helmets remove. It may well be the case that helmets make bike accidents involving banging the head safer, but they make accidents in general more likely, it seems. In short requiring their use involves risk transformation rather than risk reduction.

Another good example is safety glass. Some forms of safety glass (as I have recently found out) don't shatter: a sheet is in fact composed of two pieces of glass bonded either side of a transparent but tough plastic film, so even if the individual glass sheets break, they remain bonded to the plastic. That's all very well if you want to avoid glass shards flying all over the place in the event of a breakage. But it does mean when you want to remove the glass you can't just cover the area with some material to pick up the shards then smash it: you actually have to cut the sheet out of the frame. This is a much more dangerous job than removing ordinary glass as a number of small cuts on my fingers testify. Very minor injuries aside, though, what is interesting is that something that was created as a safety feature actually turns out to increase risk in certain situations. A good question therefore in designing any kind of safety mechanism - in mechanisms, electronics, software, finance or regulation - is:

When might the behaviour we think is unsafe actually be useful? And what will happen then if the safety mechanism prevents it?

One might even argue that collateral is functioning that way at the moment. Clearly in ordinary conditions, calling collateral reduces risk. But if the failure to post collateral actually pushes your counterparty into default, as happened to Carlyle Capital Corp, it might not be such a good idea after all. And, as JPMorganChase recently pointed out, the banking system is currently facing a systemic margin call. Hmmm....

What worked, what didn't

The Senior Supervisors Group recent document, Observations on Risk Management Practices during the Recent Market Turbulence is a more interesting read than the Financial Stability Forum interim report not least because it compares leading firms who have come through recent events relatively unscathed with those that suffered some of the largest impacts.

Some of the highlight follow, with my emphasis. First the SSG identifies four firm-wide risk management practices that differentiated performance:

Through robust dialogue among members of the senior management team (including the chief executive officer, the chief risk officer, and others at that level), business line risk owners, and control functions, firms that performed well through year-end 2007 generally shared quantitative and qualitative [risk] information more effectively across the organization.

This is pure risk culture. Firms which keep risk data to the high priesthood of risk management and a few senior business leaders are not using as much of the brainpower of their organisation as they could. Further, this kind of culture tends to go with authoritarian, vertical management styles which makes it much harder to discuss issues openly and honestly.

At firms that performed better in late 2007, management had established, before the turmoil began, rigorous internal processes requiring critical judgment and discipline in the valuation of holdings of complex or potentially illiquid securities. These firms were skeptical of rating agencies’ assessments of complex structured credit securities and consequently had developed in-house expertise to conduct independent assessments of the credit quality of assets underlying the complex securities to help value their exposures appropriately. Finally, when they reached decisions on values, they sought to use those values consistently across the firm, including for their own and their counterparties’ positions. Subsequent to the onset of the turmoil, these firms were also more likely to test their valuation estimates by selling a small percentage of relevant assets to observe a price or by looking for other clues, such as disputes over the value of collateral, to assess the accuracy of their valuations of the same or similar assets.

Clearly anyone trading complex securities should have independent expertise to value them. Equally clearly once they have been valued, that value should be used consistenly throughout the firm. What's shocking here is that some leading firms - this survey comprises eleven of the largest banks - did not have these basic control processes in place.

Those firms that avoided more significant problems through our year-end review period aligned treasury functions more closely with risk management processes, incorporating information from all businesses in global liquidity planning, including actual and contingent liquidity risk. These firms had created internal pricing mechanisms that provided incentives for individual business lines to control activities that might otherwise lead to significant balance sheet growth or unexpected reductions in capital. In particular, these firms had charged business lines appropriately for building contingent liquidity exposures to reflect the cost of obtaining liquidity in a more difficult market environment.

It so often comes down to incentive structures doesn't it? If you charge businesses for actual and contingent liquidity, including writing liquidity lines to conduits, then they will make sure the firm is paid enough for these structures. If you don't, they will be pretty much given away.

Firms that tended to avoid significant challenges through year-end 2007 typically had management information systems that assess risk positions using a number of tools that draw on differing underlying assumptions. Generally, management at the better performing firms had more adaptive (rather than static) risk measurement processes and systems that could rapidly alter underlying assumptions in risk measures to reflect current circumstances. They could quickly vary assumptions regarding characteristics such as asset correlations in risk measures and could customize forward-looking scenario analyses to incorporate management’s best sense of changing market conditions. Most importantly, managers at better performing firms relied on a wide range of measures of risk, sometimes including notional amounts of gross and net positions as well as profit and loss reporting, to gather more information and different perspectives on the same exposures. Moreover, they effectively balanced the use of quantitative rigor with qualitative assessments.

This is very interesting. If you have different tools with different assumptions, you have a reasonable handle on model risk. If you have one system, or multiple systems which all use the same assumptions, then you don't. Flexible systems are clearly important, but perhaps even more significant is the mindset of looking at risk in different ways, and being sceptical about the results of any one analysis.

The report then goes on to look at three business lines where varying practices produced different outcomes. The first, unsurprisingly, is CDO structuring, warehousing, and trading businesses. Here the key issue was whether

Internal incentives were missing or inadequately calibrated to the true risk of the exposures to the super-senior tranches of CDOs.

To be fair, I doubt anyone really got this right. What may have happened, though, is that positions that were originally acquired with the intention of selling them on became prop positions when they couldn't be sold at the mark rather than being written down until they did sell.

Next, syndication of leveraged financing loans:

Some firms worked aggressively to defend or expand market share in the syndication of leveraged financing loans, and many of those that later faced challenges in this business did not properly account for the price risk inherent in the syndicated leveraged lending pipelines.

Buying business is fine. But if you are going to pay for league table position, then at least know how much you paid for it.

Finally in conduit and SIV business we find:

Several firms did not properly recognize or control for the contingent liquidity risk in their conduit businesses or recognize the reputational risks associated with the SIV business.

For me the interesting part here is the reputational incentive. All the accounting and regulatory arguments that got these assets off balance sheet depending on the risk really passing to SIV and conduit investors. Firms that recognised that their reputational risk tolerance meant that this risk really had not been transferred clearly did better than those that pretended that the accounting and regulation followed the reality.

Pap(er) from the Financial Stability Forum

The financial stability forum working group on the crash has produced an interim report. It is an insightful document as it gives some clues about the regulator's thinking (and lack thereof). I will focus on the suggested areas of action at the back of the document: the first part discusses the causes of the crash, and that is rather old ground.

1. Supervisory framework and oversight
Capital arrangements: A resilient framework for capital requirements is central to creating appropriate capital buffers in the system and the right incentives for risk management. The implementation of Basel II, and the use of its three reinforcing pillars, is an important step to achieve this.

I read that with a heavy heart. Basel II is fatally flawed, as a number of authors have demonstrated, and just pressing on with it is tantamount to fiddling with Rome burns.

The Basel Committee will take account of the lessons from recent events and assess whether refinements to the Basel II framework are needed, including with respect to the calibration of certain aspects of the securitisation framework.

It is not just the securitisation framework. It is the entire philosophy behind and calibration of the accord. Remember that residential mortgages were one of the bigger winners in Basel II. Remember that ratings are central. Remember the bizarre behaviour of the IRB formula. The supervisors need to start again.

The turmoil has demonstrated the need for larger and more robust liquidity buffers and an internationally shared view among supervisors on sound liquidity risk management guidelines.

How about capital for liquidity risk? Or if the regulators are not willing to go that far, they will at least have to say what they mean by an adequate liquidity buffer. How would you work out how much is enough?

Firms’ managements need to act proactively in response to stress test results.

This is the hardest thing for a supervisor - forcing firms to conduct stress tests is easy. Forcing them to respond to the results of them is much harder. How exactly are they going to require firms to act without taking over management's role?

Off-balance sheet activities: Basel II strengthens incentives in the financial system to manage risks appropriately and will reduce the regulatory arbitrage that generated large off-balance-sheet risk exposures.

Basel II just creates different regulatory arbitrages, at least as currently drafted. Putting a cap on the benefit from any securitisation would be easy first step towards removing those, but the suspicion is that the supervisors are mired in attempts to revise the already complex language of Basel II without seeing the bigger picture. Off balance sheet vehicles are fine provided they are genuinely off balance sheet and there is no implicit support. The problem is that supervisors currently have a set of rules which is penal with regard to genuine risk transfer and laughably generous with regard to sham transfer. Until they can figure out which is which any attempt to revise the rules is likely to damage perfectly reasonable trades.

2. Underpinnings of the originate-to-distribute model
The underpinnings of the OTD model – including origination and underwriting standards, transparency at each stage of the securitisation process, the role and uses of credit ratings – need to be strengthened. [...]

This paragraph is motherhood-and-apple-pie. Pious hope may be necessary, but so are concrete proposals.

3. The uses and role of credit ratings
Investors, many of whom have relied inappropriately on ratings in making investment decisions, must obtain the information needed to exercise due diligence.

Investment guidelines should recognise the uncertainty around ratings an differentiate products according to their risk characteristics.

Yes, but you don't regulate many of the investors, so while this is true it does not amount to a mug of beans.

CRAs must clarify and augment the information they provide to investors on structured finance products. They should ensure that uncertainties surrounding their models and rating methodologies are made transparent. We welcome that CRAs are considering differentiating ratings of such products from corporate ratings. [...]

That is a reasonable idea but it is hard to see how to standardise model risk disclosure. The concern will be that these disclosures will turn out to be boilerplate text rather than a real attempt to analyse the sensitivity of a securities' value to the choice of model.

CRAs need to take adequate steps to address concerns about potential conflicts of interest, including concerns about their remuneration models.

Clearly. That bullet is going to be very difficult for the agencies to dodge.

4. Market transparency
Financial institutions need to improve the usability of disclosed information about risk exposures and valuations, including those related to structured products and off-balance sheet vehicles

Again, the industry has proved adept at making disclosures that meet the required standards but don't actually disclose much useful information. Just look at the memo accounts for any big investment banks. So while one can sympathise with the idea of more transparency, it will need very careful management to be effective.

Further improvements are needed in firms’ valuation methodologies and in the data that they use as inputs to their valuation processes, in particular when markets are illiquid.

You can't improve the view in a mist. Valuation in the presence of size issue, liquidity issues, or model risk is inherently uncertain. Instead the users of financial statements need to be aware of that uncertainty, and we need to develop ways of quantifying it.

5. Supervisory and regulatory responsiveness to risks
Supervisors, central banks and financial authorities - individually and collectively - need to become more effective in translating risk analysis into action.

The point above applies. It is easy to say this. What are you going to do to make sure that this action actually happens?

6. Authorities’ ability to respond to crises
Central banks’ operational frameworks must be able to supply liquidity effectively when markets and institutions are under stress. Central banks are actively investigating what lessons they can draw from recent experiences for their operational frameworks, including the capacity to provide liquidity broadly and flexibly under stressed conditions, for their communication with markets, and for the steps that might be advisable across central banks to address liquidity needs in globalised financial markets.

All of this is a little depressing. It feels as if the FSF is flailing around at roughly the level of an undergraduate seminar. We are going to need more than a bit of disclosure, a wider range of collateral at the window, and management promising on the life of their favourite pet that honestly they will take action if risk gets too big to get us out of this mess.

Update. The FT uses the FSF paper as a launchpad to discuss the broader lessons of the crunch here. Meanwhile Larry Elliott in the Guardian makes the interesting point that the FED, the ECB and the Bank cannot all be right about rates. I might get to that on Wednesday.

Wolfing down the crunch

There is a fascinating article by Martin Wolf in today's FT. As usual, let me quote selectively and comment.

[The credit crunch has] called into question the workability of securitised lending, at least in its current form. The argument for this change – one, I admit, I accepted – was that it would shift the risk of term-transformation (borrowing short to lend long) out of the fragile banking system on to the shoulders of those best able to bear it. What happened, instead, was the shifting of the risk on to the shoulders of those least able to understand it. What also occurred was a multiplication of leverage and term-transformation, not least through the banks’ “special investment vehicles”, which proved to be only notionally off balance sheet.

I would distinguish between asset-backed lending and securitised lending, but Wolf is broadly right. It has turned out that the information asymmetry problems in the ABS market are too large to be easily dealt with in some case. The lack of alignment of interests has made this worse.

Banks became more leveraged through the use of SIVs, conduits and so on, but these vehicles just allowed banks to do what they always did - take liquidity, default and term structure risk - more efficiently. The real issue is why only one of these risks, default risk, has regulatory capital assigned against it. (There is no capital charge for interest rate risk in the banking book, and no capital charge for liquidity risk.)

What, more precisely, should a central bank do when liquidity dries up in important markets? Equally, the crisis suggests that liquidity has been significantly underpriced.

Cut rates and broaden the range of collateral eligible at the window, as the FED has just done.

Does this mean that the regulatory framework for banks is fundamentally flawed?

Yes. See here, here, here and here.

What is left of the idea that we can rely on financial institutions to manage risk through their own models?

A better understanding of model risk as here, here, or here.

What, moreover, can reasonably be expected of the rating agencies?

Not a lot. Why did you ever think otherwise?

A market in US mortgages is hardly terra incognita. If banks and rating agencies got this wrong, what else must be brought into question?

It's not the market, it is the structure. Dollar yen spot is probably the most liquid asset in the world, and certainly one that is very well commented upon. Yet I can structure a dollar yen exotic option whose price is genuinely uncertain (because it is radically different depending on your modeling assumptions). There is a measure of caveat emptor here, though: why did people buy structures they did not understand?

Do you remember the lecturing by US officials, not least to the Japanese, about the importance of letting asset prices reach equilibrium and transparency enter markets as soon as possible? That, however, was in a far-off country. Now we see Hank Paulson, US Treasury secretary, trying to organise a cartel of holders of toxic securitised assets in the “superSIV”. More importantly, we see the US Treasury intervene directly in the rate-setting process on mortgages, in an attempt to shore up the housing market.

As George Monbiot pointed out in a nice article about Matt Ridley (ex chairman of Northern Rock), it is often the most vocal proponents of the open market for other people who are the most dirigist when it comes to their own business. Yes, there is a massive measure of hypocrisy here, and moreover the intervention is not well-designed: the MLEC is looking more and more dodgy, and the Bush proposal for mortgage modifications will either cover very few borrowers or make lawyers rich.

A US recession is possible.

Yep. Likely even.

What does a CRO do all day?

I'm late on the news that JP has hired a chief risk officer. (On the other hand it took JP a year to get around to it, so arguably I'm not the only one who is late.) Anyway, this brought to mind an old article of Rick Bookstaber's about what a CRO does. Bookstaber says in part:

What about the job of the risk taker? Well, a risk taker does, after all, take risk. He tries to do so intelligently, that is, he tries to put on positions that he hopes have a high return per unit of risk. But how much risk he takes and where he takes it has to be dictated by someone. You can’t just say “take risk, and good luck”.

The job of the risk manager at these firms is to convey the risk parameters to the risk takers, to define the boundaries.

I'm not sure that I buy that as the complete job description. To me it smells too much of risk measurement plus limits. Where is the risk management? Surely a CRO should be involved in advising the board on the firm's overall risk appetite, its macroscopic risk position, and capital allocation. He or she isn't just a goto person for yes/no decisions on positions the traders like: they should be proactively working with the Board to tailor the firm's risk position to optimise ROE and protect shareholders.