Charles Littrell, Executive General Manager - Australian Securitisation Forum, Sydney
I propose to take up this year’s address where I commenced last year’s speech to the ASF, by reminding ourselves of APRA’s statutory mission.
We are required to balance safety with efficiency, competition, contestability, and competitive neutrality.
We are also required, as a matter of equivalent importance, to promote stability in the Australian financial system.
In carrying out our mission, APRA much prefers to take a principles-based approach. In the case of securitisation over the past decade, APRA has found itself relying a bit too much on detailed rules. Our proposed reforms are intended to redress this issue.
As APRA has reconsidered its prudential approach to securitisation during the past year, we have sought to re-define the principles we apply in this area. Under the current prudential arrangements, the core principles are:
- an ADI must deal with the Special Purpose Vehicle and its investors on an arm’s-length basis and on market terms and conditions;
- an ADI must clearly disclose to investors the nature and limitation of its involvement in a securitisation;
- an ADI’s involvement in a securitisation must be set out in legal documentation and be limited as to time and amount;
- an ADI must not provide, or knowingly create or encourage a perception that it will provide, support to a securitisation in excess of its explicit contractual obligations.
All these principles are fine, as far as they go, but they are much closer to legal requirements than would be the case in a typical APRA prudential approach.
We propose to add the following new principles, as we reform APS 120:
- the securitisation prudential regime explicitly allows for both funding-only and capital relief structures;
- credit risks in securitisation are both clearly assigned and, for APRA-regulated entities properly capitalised;
- securitisation credit and liquidity risks are distributed in a fashion that reduces rather than increases systemic risk;
- the maturity profile of securitisations must appropriately match the maturity profile of the underlying assets or, for short-term assets, match the maturity profile of winding-down or divesting the relevant line of business; and
- securitisation is simple to understand, transparent and low risk.
From the mid-1990s to 2007, APRA and industry discovered that securitisation possesses many attractive features, which are particularly relevant to Australian home lending. These attractive features probably started with contestability: securitisation allowed non-ADI lenders, and smaller ADI lenders, to enter and compete with Australia’s major banks for home loans, across the country. This in turn substantially improved competition, and from the mid-1990s we saw a marked reduction in lending spreads on home loans.
As for efficiency, in this context efficiency is driven by allocating credit and liquidity risks where they are best borne. One could argue either way on this point.
Competitive neutrality refers to the boundary between public and private sector activity in the financial markets. Securitisation had the attractive feature in Australia of allowing a large increase in home lending, without any overt financial participation by the public sector. This is not always the case with securitisation, as the U.S. experience with Freddie Mac and Fannie Mae demonstrates.
On the safety front, up to about 2007 APRA was becoming concerned, but not yet alarmed, by some of the trends we saw in the securitisation market. Concern grew from the industry’s evident addiction to unnecessarily complex structures, and the sense that some originators were treating securitisation, combined with lenders mortgage insurance, as a kind of perpetual motion machine. Lending could be increased without the need to increase the liquidity or capital backing for this lending. There was also the concern that whatever rule APRA put in place, the industry sometimes sought corner solutions, which might technically comply with the letter, but not necessarily the spirit, of the relevant standard.
Overall, APRA’s view was that securitisation was clearly a net positive for competition and contestability, but needed to be carefully monitored for emerging trends. APRA’s supervisory work generally indicated that the potential for reckless lending through securitisation was not materialising in fact.
During the Basel II roll-out APRA updated its securitisation approach, notably moving away from pre-vetting transactions. When the global financial crisis (GFC) hit, our supervisory resources were drawn away from securitisation for about 18 months.
From around 2010, we and everyone else had a different view about the risks and benefits of securitisation, and this view had considerably darkened.
Most notably, and this is a lesson much broader than securitisation, we discovered that international money markets could shut, possibly for some time, to even the best credit propositions. Which of course makes securitisation a brittle reed upon which to base an ADI’s funding strategy, particularly any funding requiring short term rollovers.
We also discovered, thankfully in the global context rather than in Australia, that securitisation could be alloyed into a chain of transactions that supported extraordinarily rash, and sometimes fraudulent lending.
On the other hand, we discovered that despite the historic global market disruptions, by and large conventional Australian RMBS arrangements were sound, in both a liquidity and credit sense.
Moving from origination to investments, we discovered that securitisation, as one of the family of complex structured credit arrangements, could facilitate losses from investments thought to be secure, but which were in fact highly risky.
We also learned that securitisation could have more systemic implications than regulators thought was the case prior to 2008. If securitisation makes aggressive lending too easy in good times, but isn’t available to fund sound lending in adverse capital markets, then pretty clearly securitisation is pro-cyclical. This is unhelpful in a systemic sense. APRA is hopeful that its proposed reforms will reduce the pro-cyclical element in securitisation.
Then there was the surprise that all the complexity in the market, touted in various forums as innovation and completing markets, turned out to be, by and large, simply a way to conceal bad credit risks from end-investors. These investors, having learned that some complex and opaque instruments could not be trusted, panicked en masse and fled from all complex credit including unsecured lending to otherwise sound banks.
And then we discovered that banking systems reliant upon wholesale and short-term credit to fund assets, including but not limited to securitisation arrangements, could require strenuous public sector intervention to keep the banking system liquid. We also discovered that the Australian Government, should it wish to keep the Australian securitisation market at least partially open in bad times, might need to become an investor in that market.
So in summary, from say 2007 to 2011 there was a fundamental global re-assessment about the value attaching to securitisation, and much of that re-assessment was distinctly negative.
And yet, even with all these issues, it seems clear that there is nothing wrong with the basic idea of securitisation. Nothing about the GFC disproved the potential value of securitisation in theory. There was an appalling range of lessons about the dangers that attach to securitisation in practice, when the tool is used to facilitate poor quality lending, fragile funding, and opaque securities origination.
APRA’s intended new approach
Accordingly, and in common with regulators around the world, APRA proposes to reform its regulatory and supervisory approach to supervision. APRA intends to take an approach firmly grounded in the economics of the relevant risks, and one that is calculated to both support a large Australian securitisation market and discourage those practices associated with unacceptable risks.
APRA’s main goals in its proposed reforms are:
- ADIs will find it straightforward to use securitisation as a funding-only tool;
- simpler requirements for ADIs seeking capital benefits;
- the new regime incorporates the lessons from the crisis, including those specifically associated with agency risk, complexity and mismatched funding structures;
- risks become more efficiently allocated among those with more or less knowledge of the risk, and those with more or less ability to bear any losses;
- addressing systemic risks as well as individual entity risks; and
- compliance with Basel Committee requirements, with appropriate Australian adjustments.
And, as has been the case for many years, we would like to see a large and liquid securitisation market supporting competition Australian home lending.
We intend to achieve these aspirations, most obviously, by rewriting APS 120, our prudential standard on securitisation, in a way that better facilitates simple and safe securitisation, and becomes more restrictive on complex and less safe structures.
We intend to base APS 120 on the concept of funding-only securitisations. These have a simple capital relief arrangement: there isn’t any, and all capital requirements stay with the originator. But such structures have the potential to raise a great deal of term funding, which is what is needed to make the Australian banking system more liquid on a stand-alone basis.
We will also continue with a framework that accommodates capital relief securitisations, but we intend to reverse course from having industry come up with structures, to defining an allowable structure. Among other things, this structure includes a substantial skin in the game requirement.
On the investment side, APRA sees the strong benefits associated with simple structures, but complicated structured credits are too amenable to misuse. Accordingly, we propose to reduce the constraints on investing in obviously safe securitisation paper, balanced by substantial disincentives to invest in less safe and less understandable instruments. APRA’s work on the investment side of securitisation will necessarily be guided by the Basel Committee’s efforts in this area. We expect the Basel Committee will follow up its current consultations with a final rules text on risk-weighting securitisation investments, during 2014.
I will now shift gear from the history and regulatory philosophy of securitisation, to what APRA intends to reform in APS 120. Allow me to insert a caveat here that all you will see is subject to revision as we move to issuing a discussion paper, and the eventual new APS 120.
For funding-only securitisations, we consider a two-class structure makes the most sense. The A class, pari passu in credit terms and as a practical matter likely to be all AAA-rated, may contain a number of tranches including bullets and pass-throughs.
The B class is a single instrument, in which essentially the entirety of the credit risk associated with the underlying assets is concentrated, and this instrument must be held in its entirety by the originator.
The current 20 per cent holding limit on instruments will be rescinded, and we propose to allow a date-based clean-up call. We do not propose to allow other date-based calls, on the basis that this option to the issuer too easily becomes a de facto obligation on the issuer.
In sum, this arrangement will allow issuers to take a quite simple approach to raising funds against an amortising asset pool.
Let’s now move to master trusts, those vehicles which seem to possess a near-mythical attraction for some in our market.
In summary, APRA proposes to allow master trusts, but not if they are de facto covered bonds. The core issue here is that the so-called seller’s share in a master trust cannot become the equivalent of excess collateral in a covered bond vehicle. This means, among other things, that the seller share must rank at least equal to the most senior instruments in the structure.
Master trusts will allow soft bullet arrangements, but APRA will be concerned to ensure that these arrangements are not so firm that the originator is on the hook for repayment, if the asset pool is insufficient to the task.
Securitisation and capital relief
Having established the foundation for funding-only structures, it is straightforward to move to securitisation for capital relief.
In a structure with a single B class instrument, the capital relief is simply the proportion of the instrument irrevocably sold, subject to a skin in the game cap of 80 per cent capital relief, or 20 per cent minimum capital retention.
In the past year we have listened to the representations from industry that limiting securitisation to a single B class is too restrictive. Accordingly, we propose to allow multiple classes of subordinated notes which need not be pari passu. Capital relief, however, is limited to the class of subordinated notes that is most retained by the originator.
Here are two examples to illustrate the point, both presuming three credit levels within the B class. In the first example, the skin in the game floor does not apply, the minimum sold proportion is class B2 at 30 per cent, so the allowable risk-weighted asset reduction is 30 per cent.
In the second example, over 80 per cent of each subordinated class is sold, so the 80 per cent risk weighted asset reduction limit applies.
The combination of these proposals means that ADI originators in RMBS structures should be able to fund up to about 98 per cent, and capitalise up to 80 per cent, of their home lending using securitisation. A number of people have pointed out that the current arrangements in theory allow 100 per cent reductions for both funding and capital. We know this, and APRA no longer considers that such an approach is prudent, based upon lessons learned from the GFC.
Unlike the current prudential arrangements, we propose to more explicitly integrate ADI liquidity with the securitisation framework.
The Basel III liquidity arrangements split between larger ADIs, mainly banks, that must model net cash outflows and apply the Liquidity Coverage Ratio (LCR) test, and smaller ADIs, mainly mutually owned, that must meet a simpler threshold for minimum liquidity holdings, the so-called MLH test.
For ADIs subject to the LCR, cash flows associated with securitisation vehicles will figure in the net cash outflow calculation, with outflow assumed from the earliest date possible under relevant calls or similar arrangements. For smaller ADIs on the MLH liquidity arrangements, a simpler approach will apply, under which longer-term securitisation liabilities will not attract liquidity requirements, but any liabilities maturing in less than a year will need some liquid asset backing.
ADIs will be allowed to trade in Class A instruments, including their own instruments. It is possible for ADIs to take advantage of this arrangement to create market expectations that they are issuing hard bullets in a funding-only securitisation. APRA also forms market expectations, and should we form the view that an ADI is gaming the trading in its own paper, the ability to do so might become limited.
For banks establishing a Committed Liquidity Facility (CLF) with the RBA, the all reasonable steps test will apply to securitisation, in the same way this expectation applies to banks issuing unsecured term liabilities. That is, banks with substantial home loan pools are in a good position to issue funding-only or for that matter capital relief securitisations, and they will be expected to do so, in order to further increase the tenor of their wholesale liabilities.
APRA has considered many issues associated with securitisation warehouses. Our inclination is to allow originators and warehouse providers to continue with warehouses, more or less with the current capital effects, provided that the underlying assets are term funded within a year. Past one year, such arrangements will be treated as a whole loan sale by the originator to the warehouse provider.
Most of APRA’s securitisation arrangements focus upon issuers. On the other side of the balance sheet, exposures to securitised assets may comprise a material part of some ADI assets. As noted previously, the Basel Committee is expected to issue an additional standard in this area during 2014. APRA’s reformed APS 120 will at least implement the Basel Committee’s minimum requirements.
One lesson APRA has taken from the GFC is that senior paper in a simple securitisation arrangement is likely to be sound paper, and we are comfortable that ADIs could hold such assets, under the relevant risk-weightings flowing from the Basel Committee.
By contrast, APRA’s tolerance for subordinated exposures, and for apparently senior instruments with underlying complexity such as CDO-squareds, has greatly reduced. Accordingly, we propose a simple rule: ADIs holding the most senior class in conventional securitisations receive a Basel risk-weight. Our current thinking is that holdings in any subordinated securitisation arrangement, or any holdings in instruments that look like CDO-squareds, would attract a CET1 deduction. In any event, holdings in another ADI’s B notes would attract a CET1 deduction.
APRA is more comfortable that less levered investors, such as super funds and most insurance companies, might prudently hold the junior or complex securitisation classes. Therefore we do not propose to apply the CET1 deduction rule to our other regulated industries.
This proposed approach is rather firmer than has been taken in some other jurisdictions. It is consistent with our view that simple securitisation structures are likely to be valuable, but this value evaporates under complex structured credit arrangements.
Other issues in brief
I will touch briefly on a few other matters, which will be included in our eventual consultation paper.
We have had a number of representations on multi-seller structures. There are a number of problems here, starting with the quite limited take-up of such structures in Australia. To make multi-seller structures work, the skin in the game requirements either become very complex, or they no longer bind upon individual originators. APRA is open for suggestions on how we might solve this conundrum, but in the absence of a solution, the new APS 120 is unlikely to accommodate multi-seller structures. Any solution in this context commences with the need to clearly link each originator’s economic outcome with the credit results from that originator’s assets in any joint pool.
APRA proposes to allow asset-backed commercial paper, most likely for funding only, and with a particular caveat. That caveat is that any originator raising short-term funds through securitisation vehicles must be able to demonstrate that it can carry the portfolio through a two-year market closure. As a practical matter, this probably means a committed line of credit from a sound third-party institution. Alternatively, the ADI may be able to demonstrate that it could run down the underlying assets faster than the securitisation liabilities. In this case, however, APRA will need to be convinced that the underlying assets do not comprise customer relationships, where the customer would be left in a credit crunch through the run-down. This means that more or less conventional banking assets such as credit cards will not be acceptable under the run-down test, and must instead meet the two-year market shutdown test.
For the past five years, APRA has operated without pre-approving securitisations. We propose to continue this practice.
The proposed new APS 120 will empower ADIs to issue clearly compliant funding-only and capital relief securitisations. Any ADI looking to push the boundaries from what is meant to be a simple compliance regime does so at its own risk. Risk in this context includes, without limitation, APRA disallowing any claimed capital relief, or APRA limiting or ending an ADI’s ability to originate securitisations.
Moving on to another point, under the covered bond legislation, ADIs face an eight per cent statutory limit on the degree to which they can pledge assets. There are cognate arguments as to why APRA might place a similar but larger limit on the degree to which an ADI may encumber assets for securitisation. APRA does not propose to take such an approach at this time, instead relying on the more general principle that ADIs must prudently diversify their funding sources. APRA may review this position for individual ADIs or the industry in the future, should we perceive that asset encumbrance is materially affecting issues such as depositor preference.
Finally, the current APS 120 allows for capital relief for some derivatives transactions. This little-used provision makes some sense when considering a single ADI in isolation, but considerably less sense at the systemic level. In essence, an uninformed risk taker is assuming credit risk from the informed originator. Accordingly, APRA proposes to remove this provision for those ADIs holding B class instruments and for synthetic securitisations.
APRA’s systemic and individual ADI aspirations
Our aspiration for securitisation under the reformed prudential arrangements is that Australia will end up with a large and active funding-only market. This will materially assist Australia’s banking sector to correct its current imbalance between long-term, retail domestic assets, and a funding side overly reliant upon the kindness of strangers. By integrating the liquidity side of APS 120 with APRA’s prudential liquidity requirements, we will better insulate Australia from future global or domestic funding market failures.
We are hopeful that the new arrangements also allow a substantial market in capital relief securitisations, under arrangements that much more explicitly transfer credit risk from the originator to other investors. It remains to be seen how this market will develop.
In any event, we consider that these arrangements will maintain robust incentives for originators to maintain reasonable lending quality.
We seek to avoid any build-up of toxic structured credit assets on the Australian banking system’s balance sheet. Australia largely dodged this bullet during the GFC, and we need to ensure that the lessons learned are captured in APRA’s prudential framework.
Finally, as with international perceptions of the Australian banking system more generally, we want international investors to consider the Australian securitisation market a place where they can easily make safe and simple investments, and are most unlikely to be subject to the undesirable practices revealed elsewhere in the world. Some future panic might better distinguish between quality in national securitisation markets. Neither APRA nor anybody else can guarantee that the Australian securitisation market won’t close. The arrangements we are proposing, however, should go some way towards ensuring that we are the market that closes last, and re-opens first.
APRA’s aspirations for individual ADIs are similar to our systemic aspirations. An ADI can view securitisation as a useful funding tool, possibly as a way to lay off credit risk, not a threat to lending discipline, and an element in the ADI’s overall liquidity arrangements.
Where to from here
APRA intends to issue a discussion paper on our proposed securitisation reforms, in the near future. With the recent change of government, some of the necessary administrative arrangements to address regulatory reform are under discussion. When these discussions resolve, we intend to issue our paper.
As with other APRA reforms, we intend to seek feedback from interested stakeholders, and refine our proposals. In any event, we do not expect the reforms to apply until 2015.
In closing, I summarise APRA’s aspirations for securitisation, which I am confident that you share.
First, we want to create a framework that allows for a large, useful, and safe Australian securitisation market.
Second, we need to take on board the lessons of recent years, in order to protect that market.
Thank you for your attention.