De-risking the banking system should be a good thing. Retail banking, deposit taking, lending lots of money to ordinary people to do ordinary things. No messing with the wholesale markets.
But when risk goes into the unregulated, free-to-fail world of hedge funds, it suddenly looks like a much scarier trade an image problem which is leading regulators to think again about how banks can get capital relief.
Regulatory capital securitisations usually mean a bank has bought protection on part of an asset portfolio, allowing it to hold less capital against the portfolio. They are called securitisations because they use similar tranching technology. Losses are allocated sequentially, and banks can choose which slice of risk to sell off. Typically this could be losses between around 1% and 8%.
However, no securities are created from this form of securitisation the usual instrument is a single bilateral CDS contract.
The problem is calculating how much risk is transferred, and working out the cost of doing so. Risk transfer is a good trade for the bank doing it if it can balance the margin thrown off by the assets with the cost of the CDS premium to get capital relief, and the reduction in the capital cost of holding the assets.
Jean-Baptiste Thiery, European head of financial institutions securitisation at Natixis, says that given that synthetic capital relief trades usually transfer the equity or mezzanine tranche, investors that want to do them are hedge funds or private equity, and not banks or asset managers.
"Sometimes, given market conditions, they have return targets above a banks cost of equity, so it can be hard to find investors at the right price to take on risk transfer," he says.
INGs global head of structured securitised finance, Amador Malnero, agrees: "Through the crisis investors that might have bought B or BB pieces started demanding higher returns. One of the issues we have today is the gap between what investors want to receive and what banks want to pay."
Bringing up Basel
If there is a gap between investor expectations and the cost banks will pay for protection on asset portfolios, new regulations, based on the Basel III framework, should change all that.
"Banks can afford to pay to get these done, because the benefits of relief will be so much greater than in the past," says Daniel Watkins of Green Street Capital, an advisory firm active in this area. "Certain assets are becoming more capital-intensive, banks need to hold more capital, and will need to pay more for capital. Its a three-pronged cost spiral."
Malnero says: "Only banks that really needed the capital did these trades during the crisis and since then, but if Basel III pushes up the average cost of capital, it could make this business more attractive."
The new international framework doesnt alter many of the basic tools for calculating risk weighting of assets, and hence capital costs. RWAs will still be the foundation of regulatory capital ratios (although the mooted leverage ratio will act as a backstop), and banks will still be able to model their own risk exposures under the Advanced Internal Ratings Based (IRB) approach.
This means the idea of de-risking a portfolio by selling off exposure to a junior tranche still works. The portfolios where banks look to save capital might change though. New risk-weightings for certain asset types that have attracted political and regulatory ire will change the incentives for doing reg cap deals.Re-securitisations and OTC derivatives, which have attracted plenty of blame since the credit crisis, will be re-weighted.
Romain Brive, head of capital structuring at Natixis, says: "There are some portfolios that will be expensive under Basel III OTC derivatives or re-securitisations for example where capital relief trades can do well, but its not an approach that works systemically for every portfolio."
Banks that are sitting on legacy ABS assets, rather than re-securitisations, would be well advised to sit tight. Green Streets Watkins explains that selling tranched equity or mezzanine risk on a securitisation book would mean the retained senior tranche would get counted as a re-securitisation and suffer a 1,250% risk-weighting.
"Finding investors to buy the equity would also mean crystallising the current market value of the legacy assets, which many banks arent keen to do," he adds.
INGs Malnero says picking an asset pool will remain a trade-off between efficiency of tranching and capital benefit. As rating agency and capital models need to operate on conservative assumptions, lumpy portfolios with large assets may need a lot of subordination to reach a safe level. This means a bank wanting capital relief needs to buy protection on a large portion of the portfolio.
For granular assets, the law of large numbers means modelling the probability of default and loss-given default in the portfolio is much more predictable, meaning banks can get away with buying protection on a much smaller slice.
Malnero says: "The tranching works well on residential mortgages and other highly granular assets, but residential mortgages dont consume a lot of capital. SME loans and traditional corporate loans are probably in the sweet spot."
While Basel III in its draft form (CRD IV in the European Union) presents an opportunity for the capital relief trade, the FSA has other ideas, and even in other jurisdictions, regulatory discretion is likely to loom large.
Malnero says the Dutch regulator didnt require ex ante approval for banks to claim capital relief pre-crisis. This was true for most regulators banks were assumed to be following the rules.
Green Streets Watkins says this could change. "Regulators wont want to sign off on large numbers of these trades," he says. "Theyll take a hands-on approach and make sure theyre happy and comfortable with each one. Itll most likely be a discretionary view, not an absolute limit."
The FSA is not content with its powers of discretionary approval. The regulator wants to get third party proof that banks have transferred enough risk before allowing a bank to claim capital relief. This means banks must get a third party rating on retained exposures, and use the Standardised Approach to calculating risk weighting instead of their own models.
This is likely to hurt the economics of capital relief securitisation for UK banks. Marginal trades will become inefficient because of the cost of getting a rating, which would be particularly pronounced for complex or hard-to-value portfolios.
Using Advanced IRB (the banks own models, signed off by the regulator) to calculate how much protection to buy is also usually more advantageous than the Standardised Approach. Whether this is because the Standardised Approach is overly cautious or because banks are gaming their regulatory capital calculations is an open question.
Market participants questioned the logic of a return to ratings.
One member of the FSAs Securitisation Standing Group (a discussion forum for market participants and regulators) said: "It is somewhat puzzling that the FSA is handing the pen back to ratings agencies as this goes against the regulatory trend elsewhere where the focus is on due diligence and firms own internal analyses."
Others questioned why the FSA did not simply recalibrate the models banks used to calculate regulatory capital requirements.
Andrew Bell, an FSA official, told the Standing Group: "There are benefits in recalibrating the Supervisory Formula Method (the approach which relies on bank models). However, this would be a time-consuming process, as it would need to go through the Basel process and feed through to the CRD."
Like many financial instruments, capital relief securitisation or risk transfer can be used for good or ill.
One anonymous market participant says: "Bankers working on capital efficiency are some of the smartest people in the capital markets. But they are fundamentally evil their job is just to arbitrage economic capital and regulatory capital."
This may be true some of the time, but mostly it is not, and it will become less so as Basel III comes in.
Attracting capital from outside the banking system to fund specific asset portfolios can be a good and wholesome thing. Banks need more equity, and it has to come from somewhere.
If it comes from capital providers that can value assets that no one else wants to, that is fine. If it comes attached to specific assets, maybe this is a better idea than a rights issue.
Bank of America shareholders, even Warren Buffett, have at best a hazy idea of what sits on its $2.2tr balance sheet. The buyers of risk transfer trades give the bank equity, but with a transparency that shareholders can only dream of.
Richard Robb, a partner in Christofferson Robb, a hedge fund that started investing in capital relief deals in 2002, is in no doubt. "Risk transfer securitisation trades are actually one of the most wholesome things going on," he says. "If you dont believe the formulae, then the best way to deal with risk in the banking system is to decentralise."He adds: "Get the risk out of the banking system and into end investors. Instead of banks going under, you get small losses at sovereign wealth funds, pension funds, asset managers and so on.