Friday, June 15, 2012

Prime collateralized securities and the safe asset designation

As part of the sell-side's push to put a new label on structure finance securities, they rolled out a list of safe assets that could be collateral for a structured finance security that receives the Prime Collateralized Security label.

Included on this list are residential mortgages.

Had this label existed 5 years ago, presumably securities backed by Irish or Spanish mortgages would have received this label.  How useful would that label be today on these securities?

The label is not a good substitute for loan-level disclosure on an observable event basis.


Anonymous said...

if you look at most Irish and Spanish mortgages within securitisations they are performing well, certainly to a higher standard than would be expected if you look at house prices and unemployment rates in those geographies.

The point of a securitisation is to put a defensive structure that can mitigate losses around an asset pool, so even with increased defaults and losses most bondholders would be made whole. Obviously this depends on what scenarios you run.

Loan level disclosure is coming as well, and is certainly available for most new issues. This is indeed not a substitute for that and noone has said it is.

Richard Field said...

The loan level disclosure that is coming will be completely worthless unless it features observable event based reporting.

Every time there is an activity with the underlying collateral, like a payment, it should be reported before the beginning of the next business day.

It is only with this type of reporting that market participants have access to current information.

To date, the efforts in Europe to provide loan-level reporting have been associated with once per month or less frequent reporting. Once per month is the same reporting frequency as provided by sub-prime mortgage backed securities and these earned the nickname opaque, toxic securities.

It is only with observable event based reporting that opacity is eliminated from structured finance.

Anonymous said...

most mortgage lenders have a limited range of payment options in terms of days of the month anyway, but there are no lenders who could currently turn around the payment information on 250,000 loans in 24hrs, that is not practical.

borrowers only pay once per month so why would you need more data to show they haven't paid when they were expected to pay.

also sub-prime, toxic RMBS are US products, check the loss experience to date on European/Australian RMSB and you will find it is in the single basis points range,and most of that is to do with events such as Lehmans bankruptcy (i.e. nothing to do with the mortgage, so outside of the remit of PCS)

Richard Field said...

Actually, the way mortgage loan databases are set up at banks is to track observable event based data.

To confirm this for yourself, try going on to the site that tracks your credit card. You will see that it has transactions through the close of business yesterday and maybe even some from today.

There is no reason that a mortgage loan database would be set up differently when it comes to tracking observable events than a credit card database and it isn't.

So, your statement about providing the data is simply false. It is actually easier for a bank to provide data on an observable event basis than to report it once per month.

As for why an investor would like to have current information on the underlying collateral performance, ask yourself which is easier to value: the contents of a brown paper bag or the contents of a clear plastic bag?

A brown paper bag is the physical equivalent of a structured finance security that provides once per month or less frequent reporting.

A clear plastic bag is the physical equivalent of a structured finance security that provides observable event based reporting.

Which is easier to value? My answer is the clear plastic bag.

With the brown paper bag you are blindly guessing as to the current performance of the underlying collateral.

There is no reason to believe that historic loss experience will be repeated. Australia's has been experiencing a real estate bubble for years. This suggests that future performance is not likely to be as good as past performance.

Without observable event based data, how is an investor suppose to determine which of the Australian RMBS to buy should performance decline?

Anonymous said...

ok point taken on the speed that payment systems are able to show cashflows, but i question how useful that is in the context of the PCS initiatives markets.

Yes clearly historic performance is not a guide to the future, and when rates go up there will clearly be pressure on households with an increased payment burden, but that doesn't get away from the point that by far and away the majority of RMBS in Europe and Australia is Prime (and no we dont ahve a FICO based definition of what that is), and performance has been incredibly steady through the last 5 years turbulence.

Invetsors in this space do not need real time data. My stress modelling on LTVs and defaults on a bond by bond basis shows that you would need to see defaults increase in the UK Prime space from something equivalent to 0.4% to 25% and at the same time you would need house price declines to the tune of 80% from current before you would take a loss at AAA level, and that is using conservative assumptions (zero excess spread, all HP declines and defaults happening today etc etc). This is just not going to happen and if it does then most banks will have gone bust long long ago.

essentially what you are advocating is probably useful in the toxic world of US Subprime, but in Europe we never got to the simply appalling level of credit underwriting, and structuring that that entailed.

PCS is trying to differentiate between these different sectors by creating a standard for collateral that has not become stressed yet and in all probability never will to the tune fo what you seem to be assuming.

Richard Field said...

"Investors in this space do not need real time data" is a statement of your opinion.

First, I am not arguing for real time data. I am arguing for observable event based reporting. Only those loans that have an activity are reported on. The report is made available to investors before the beginning of the business day following the business day the activity occurred on.

Second, given that the servicers can easily provide observable event based reporting, what benefit is there to investors from not having observable event based reporting?

There isn't one.

Regardless of how frequently (once a day, once a week or once a month) an investor chooses to look at the underlying collateral performance, with observable event based reporting they are always looking at the current performance of the underlying collateral.

I can easily imagine that the investors in the highest risk tranches of a structured finance security would want to look at the collateral performance once per day.

AAA-rated investors might want to look once a week or less frequently.

There is no benefit to the AAA-rated investor in denying the investors in the riskier tranches access to information that meets the riskier investors preferences.

PCS is the equivalent of putting a label on a skunk and calling it Pepe le Pew. The label tells an investor absolutely nothing about the current quality or performance of the assets underlying a structured finance security.

PCS is supported solely by the sell-side. It is being offered as a substitute for disclosure and providing all market participants with observable event based reporting on the underlying collateral.

Why is the sell-side pushing what is effectively a meaningless label and not pushing for observable event based reporting?

Anonymous said...

Pcs is not only supported by sell side, it has support cross industry from investors, sell side, regulators and central bankers. If you were at the global and conference in brussels last week you would have seen this.

I don't think investors in prime RMBS deals would be looking at collateral performance as often as you state, but I am not saying that relevant timely performance data is not relevant, I just think that given performance levels vs credit enhancement levels and excess spread levels investors are not that focussed. Our current Market is driven much more by liquidity, bid offer spreads, Eurozone, growth data etc. Relevant performance data will become more relevant when fundamental performance is again the driver of spreads, but actual losses for prime securities are a long way from bEing a current issue. Remember as well that most subordinated prime securities have matured and been replaced by retained bonds sitting with the originator.

PCS is about getting invEstors buying high quality mortgages again. Your sporting would donthe same but for sub/mezz buyers. They are not mutually exclusive. But be aware that PCS is not trying to be a proxy for a rating, it only references the asset quality at launch.

Richard Field said...

PCS is a sell-side led initiative. They ponied up 3+ million euros to fund a secretariat. I had an earlier post on this linking to a Reuters' article.

Next, the ECB is desperate to unload the structured finance securities on its balance sheet and reduce its risk. This simple fact undermines any credibility that its endorsement might carry.

Next, I am not surprised to hear the claim that regulators support/endorse PCS. Like the ECB, their endorsement is undermined as they have an incentive that undermines their credibility. In the case of the bank regulators, they are desperate to find funding for their banks so they can unload the retained structured finance securities.

Finally, I would be highly surprised to see a major campaign by investors expressing support for a label. A label does not relieve them of the legal liability to 'know what they own', which they know is impossible in the absence of observable event based reporting, or the need to do due diligence.

Actually, the PCS label is not just about the asset quality at launch. It specifically makes a representation as to the quality of disclosure that will be provided by the deal.

Unless the deal provides observable event based reporting, a PCS label should indicate it is unsafe.

I know you don't think that investors will look at the data that often, but have you considered that if the data were available there are industry experts like pricing services that would?

I stand by my statement that there is no benefit to investors from not having observable event based reporting.