Tuesday, August 2, 2011

Complexity, concentration and contagion solved by the FDR Framework

Since before the credit crisis, your humble blogger has said and demonstrated that the issues of complexity, concentration and contagion in the financial system are most effectively and efficiently addressed and solved with the simple mechanism of disclosure.

Regular readers are familiar with how disclosure works in the context of the FDR Framework.  Under this framework, governments are responsible for ensuring that market participants have access to all the useful, relevant information in an appropriate, timely manner.  Market participants, since they are responsible for absorbing losses on investments under the principle of caveat emptor, have an incentive to use this data to analyze the risk of any investment and adjust both the price and amount of their exposure based on the results of this analysis.

Recently, the Bank of England's Andrew Haldane delivered a speech on haircuts in the repurchase agreement markets.  This speech was a prelude to the forthcoming publication of an article he co-authored titled "Complexity, Concentration and Contagion".

Like Gary Gorton and Andrew Metrick before them, Haldane and his co-authors develop a theory of how secured funding markets work and a model.  As with Gorton and Metrick, Haldane and his co-authors' theory and model are inferior in every respect to the parsimonious FDR Framework.

Gorton and Metrick observed what they called a run on the repo.  Haldane and his co-authors observe that the amount of collateral required for a repurchase agreement jumped significantly from before the credit crisis to the peak of the credit crisis.
Table 1 demonstrates this pattern [for prime counter-parties]. It compares haircuts on a range of financial instruments used to back borrowing – so-called securities financing. They are shown on two dates, before (June 2007) and after (June 2009) the financial crisis. Haircuts rose by up to 90 percentage points in the space of these two years, as the scorching pre-crisis summer gave way to a frozen crisis winter. In other words, haircuts exhibited a rather dramatic pro-cyclicality over the course of the crisis.

Table 1: Typical haircut on term securities financing transactions (per cent)
                                                                             June 2007      June 2009
Medium-term G7 government bonds                         0                     1
Medium-term US agencies                                        1                     2
AAA-rated prime MBS                                            4                    10
Asset-backed securities                                           10                    25
AAA-rated structured products                               10                  100
AAA- and AA-rated investment grade bonds           1                     8
High-yield bonds                                                      8                    15
G7 countries equity                                                10                    15
Source: Committee on the Global Financial System (2010).
This haircut cycle played an important causal role in the crisis. Secured financing became an increasingly important source of credit in both bank and non-bank markets over the past decade. In the US, the repo market financed roughly half of the growth in investment banks’ balance sheets between 2002 and 2007. In the UK, the securitisation market trebled in size over the same period. Those were the heady days of summer. Since then the US repo market has shrunk by 40%, while the UK securitisation market remains frozen.
Why did this increase occur?  Was it predictable? (hint:  the increase was predictable and it is a matter of record that using the FDR Framework your humble blogger predicted it along with the credit crisis)

As discussed in a previous post, Gorton and Metrick acknowledge that they and their theory of how secured funding markets work using informationally insensitive debt are unable to answer these basic questions.  They wrote [link in previous post]
The reason that this shock occurred in August 2007 – as opposed to any other month of 2007 – is perhaps unknowable. We hypothesize that the market slowly became aware of the risks associated with the subprime market, which then led to doubts about repo collateral and bank solvency. At some point – August 2007 in this telling – a critical mass of such fears led to the first run on repo, with lenders no longer willing to provide short-term finance at historical spreads and haircuts.
According to an article by Larry Elliott, the Economics Editor for The Guardian,
On the face of it, there was nothing especially memorable about August 9 2007.  
... It was, however, the day the world changed. As far as the financial markets are concerned, August 9 2007 has all the resonance of August 4 1914. It marks the cut-off point between "an Edwardian summer" of prosperity and tranquillity and the trench warfare of the credit crunch - the failed banks, the petrified markets, the property markets blown to pieces by a shortage of credit. 
On that day, the European Central Bank and the US Federal Reserve injected $90bn (£45bn) into jittery financial markets.
What happened on August 9, 2007?

According to a report by BBC News,

Investment bank BNP Paribas tells investors they will not be able to take money out of two of its funds because it cannot value the assets in them.
Specifically, it could not value the subprime mortgage backed securities in these funds.

Why could it not value these securities?

As was disclosed by the rating agencies in their testimony before the US Congress a month later, these securities could not be valued because there was no access to the data needed to monitor them and to make timely rating changes.

As all market participants know, ratings are just a form of valuing a security.

Under the FDR Framework, what happened in August 2007 is not the unknowable event Gorton and Metrick claim, but rather an easily predicted one.  It was the month in which BNP Paribas announced that all the useful, relevant information about the subprime mortgage backed securities was not available in an appropriate, timely manner and therefore the securities could not be valued.


Please refer back to table 1.  In particular, look at what happened to the haircut on structured finance products (highlighted).  Under the FDR Framework, it is not surprising that the haircut on structured finance products increased significantly.  If you cannot value them, it is irresponsible to lend aggressively against them even to prime counter-parties.


But this only tells part of the reason why haircuts would increase.  The other issue is the notion of a prime counter-party.  At the beginning of the credit crisis, market participants did not have all the useful, relevant information they needed in an appropriate, timely manner to tell if any prime counter-party was solvent or not.  This dramatically increased the probability of ending up with the collateral. 


This too is predicted and easily explained under the FDR Framework.


In his speech, Mr. Haldane does not attempt to address the questions of what caused the increase in the haircuts and was it predictable, but instead introduces a model and how the model can be used to test different policies options.

The model's predictions do not measure up to the accuracy and predictive value of those predictions generated under the FDR Framework [and stated in this blog].

Take for example the prediction that if there was a 20% haircut on secured financings there would be a very low probability of a liquidity crisis.  Had the model been used prior to the credit crisis in 2007, with a 100% probability there would still have been a liquidity crisis in the secured funding market.

What needs to be better appreciated is that market participants could not value the collateral (structured finance securities) or evaluate the solvency of the borrower and therefore elected not to make a "secured" loan.  This froze the market.

As regular readers know, I have been using the FDR Framework since the beginning of the credit crisis to make predictions about the success or failure of different policy options (see the posts on Ireland for example).  As regular readers know, the accuracy of these predictions has been extraordinarily high.

Once again, I am offering my services to Mr. Haldane and the Financial Policy Committee when it comes to evaluating different policy options.

No comments: