Thursday, October 15, 2009

How the Basel II capital cliff begets resecuritisation P

Basel II banking regulations are sooooo boring.

But Basel II’s effects on securitisation are fascinating, right?

Let’s begin.

Rating agency Fitch is guiding us through the Basel II regulations, currently being rolled out around the world, and their impact on securitisations — the financial exotica that are RMBS, ABS, CDOs and the like.

Here’s what Fitch says on the matter:

A core objective of Basel II is to close the regulatory capital arbitrage that exists under Basel I. Basel I’s crude approach to measuring risk gives banks an unintended capital incentive to securitise high quality, low‐yielding assets and retain low quality, high‐yielding assets, since the amount of capital required against each is identical. This tactic of securitising high quality assets and retaining riskier exposures, including subordinate tranches, decreases Basel I capital requirements without necessarily reducing economic risk exposure.
This is something we’ve heard before. Specifically, the idea that the modified Basel II rules will effectively kill the market for complex products, since they require more capital — in some cases up to 3.5 times more — to be held for some securitisations. In other words, banks will no longer want to hold onto the riskiest tranches of securitised stuff, and without banks to hold the mezzanine tranches, they can’t sell the less risky tranches to others.

Here’s what Fitch has to say about the capital weightings for different shades of securitisations:

Finally, both the standardised and IRB approaches feature a “cliff” in securitisation capital charges between investment grade and non‐investment grade exposures. This ramp‐up in Basel II charges is broadly comparable to the profile of average annual structured finance impairment rates, which also increase markedly when crossing below the investment grade threshold (see Chart 1).

In practice, that cliff effect looks like this:

Basel cliff effect - Fitch

What’s interesting, however, is that Fitch doesn’t think this will lead to fewer securitisations but to more:

This cliff effect is also one of the reasons behind the recent growth in re‐securitisation activity as banks are repackaging distressed or heavily downgraded securities to achieve higher ratings on the resulting senior tranche, which in turn translate into lower capital charges if retaining the senior tranche.

Let’s rewind a bit here.

The point of Basel II is to neutralise capital arbitrage risk:

Basel II is designed to neutralise this arbitrage opportunity by aligning regulatory capital charges more closely with economic risk, thus reducing capital incentives both to securitise high quality assets and to retain subordinate tranches. By promoting greater risk‐sensitivity, Basel II moves closer to achieving capital neutrality (or capital charges that are identical for both an unsecuritised pool of assets and a securitisation of these same assets). Since the securitisation process typically changes the form but not the overall amount of risk within a pool of assets, the total capital charges on the unsecuritised pool should equal the total charges across the full securitisation structure.

That should result in the kind of regulatory framework the Basel Committee is aiming for — one in which banks decide to securitise assets based on actual economic interests rather than pure capital arbitrage.

So does Basel II succeed in creating equal charges for unsecuritised and securitised assets?

Fitch has helpfully calculated a few capital requirements on sample securitisations under Basel II. Here are the results for the Internal Ratings-Based methodology (where banks are allowed to use their own empirical models to estimate probabilities of default) and the standardised approach (using external credit assessments - i.e. ratings agencies). The blue ‘underlying assets’ bars are basically the unsecuritised versions of banks’ assets:

In the CMBS and Credit Card ABS examples using the standardised approach, the unsecuritised assets require more capital than their securitised counterparts. Now, Fitch is very clear that these are illustrative examples only and that readers should not generalise based on these results. The reason for that, Fitch says, is because capital weightings for securitisations under Basel II are extremely sensitive to a variety of factors.

For instance:

Simple scenario testing illustrates the dynamic nature of Basel II … For the sample CMBS transaction in this study, assume instead that (1) the PD [estimates of probability of default] is 0.60% instead of 0.75% (eg the IRB bank uses different empirical data or default risk models to estimate PD) and; (2) the four senior tranches (which together account for over 90% of the structure’s total notional exposure) maintain their credit ratings but that the two lowest tranches (which only account for about 7% of the deal’s total exposure) incur a two to three notch downgrade (ie the ‘BBB’ tranche is downgraded to ‘BB’; the ‘BB’ tranche falls to ‘B+’).

The rather stark disparity in capital charges in the base case (ie 5.4% capital if unsecuritised versus 2.5% if securitised) is now perfectly aligned (ie 5% capital on both an unsecuritised and securitised basis) in this new scenario. A relatively small shift in risk measures (ie decreasing PD by 0.15% and a few notch downgrades affecting only 7% of the capital structure) has fully neutralised what was previously a two‐fold difference between the unsecuritised and securitised Basel II charges.

What should be becoming clear here is that the structure of the resecuritisation is a crucial element in deciding capital weightings under Basel II.

A subtle change in the structure — default assumptions, credit enhancement, etc. — can result in a vastly different capital requirement:

Given the “cliff” in Basel II securitisation charges across the investment grade threshold and the acceleration in charges when moving down the ratings scale, it is apparent that relatively subtle differences in capital structure can have a significant impact on the resulting Basel II charges. Marginal changes in the portion of the capital structure rated below ‘BBB‐’ (and, particularly, the portion rated below ‘BB‐’ and requiring a 100% capital charge) have a magnified impact on overall Basel II charges for the transaction.

Basel II’s risk‐sensitivity goes a long way towards closing Basel I’s regulatory capital arbitrage, which was driven in large part by crude risk‐weighting approaches resulting in the same blanket capital charges applied to differing risk exposures. Under Basel II, however, the same or similar risk exposures could in many instances face different capital charges, potentially creating new forms of regulatory arbitrage in which banks hold a given risk exposure based on the methodology and form generating the lowest capital requirements. This potential arbitrage stems partly from Basel II’s flexibility in providing a range of calculation approaches tailored to different banks, risk quantification methodologies, risk measures, asset types and forms of risk exposure.

In other words, more of the same kind of behaviour in which banks have been engaged for the past decade or so.

No comments: