Thursday, March 23, 2017

The Art of (Illiquid) Securities Pricing

As you all know, the financial meltdown was caused by some part faulty-product (mortgages, RMBS, CDOs) and some part market-panic itself and its influence on certain other products (auction rate securities, SIVs) and market mechanics (pricing, rating).

Faulty products are not new: the world is awash with faulty products.  But we need buyers for them, and to encourage buyers we need forums (e.g. securitization) and mechanisms (e.g. ratings) that would induce buyers and give them comfort that the faulty products weren't, err, all that bad.

Well that's a long and old story.  But where we're going today is that many of the mechanics that went awry, and needn't have, have not been fixed.

Back in 2008, we had majestic moments of illiquidity, which spurred quotes like this one, from a conversation among AIG employees:
“we can’t mark any of our positions [to market price], and obviously that’s what saves us having this enormous mark to market. If we start buying the physical bonds back then any accountant is going to turn around and say, well, John, you know you traded at 90, you must be able to mark your bonds then.”
Since the crisis, the SEC has ramped up its investigations into pricing issues, and in 2013 set in motion three initiatives (the Financial Reporting and Audit Task Force; the Microcap Fraud Task Force; and the Center for Risk and Quantitative Analytics).  There has been a steady and growing stream of findings of asset valuation mismanagement.  Some hedge funds have been shut down. (A list of issues here.)

But there is much to be done, if the recent dispute between a Canadian pension fund and a US hedge fund is anything to go by.

We have written about that dispute in detail here and here, but a transcript was released as part of discovery in the matter that depicts just how tricky and error-prone our pricing systems are as soon as there is any level of illiquidity.  Before we get to the transcript, here's a brief picture of the issue at play, per the pension fund's (original) allegations.


  • Pension fund requested a full redemption of its investment in Saba’s hedge fund 
  • Prior to redemption, Saba marked down its valuation of one issuer’s corporate bonds (a relatively illiquid issuer), lowering the fund’s NAV and the amount to be returned to redeeming investors
  • Saba altered its valuation methodology to mark down the bonds issued by The McClatchy Company (“MNI”): it applied a bids-wanted-in-competition (BWIC) approach instead of relying, as usual, on its external pricing sources 

    • Saba had made sales of MNI bonds in March 2015 at prices from 58% to 60% (of par)
    • 3/31/15 mark used for redemption, based on all-or-none $50 mm BWIC: 31% 
    •  Saba later made sales of MNI bonds in April 2015 at prices from 53.75% to 55.75%
  • After the redemption was completed, Saba resumed its prior valuation methodology for MNI bonds, marking them back up

With that background, here is a concise depiction (provided here by Bloomberg columnist Matt Levine) of the hedge fund's chats with its pricing providers in 2015, demonstrating just how much the pricing process is based on art, rather than science.

  • Saba Capital's Weinstein: Z, where would you bid a few mm of the 29s with or without 5yr cds? ... 
  • Trader: most likely below where you care. 50- 2mm 
  • Weinstein: Yes, that is low I think. 
  • Trader: where would u bid? 
  • Weinstein: Who knows. See it quoted much higher. Actually you should change your 65/66 quote I guess. 
  • Trader: im happy to reflect any market you would like me to make 
  • Trader: i have no position 
  • Trader: and quote it only 
  • Trader: but thats the discount i would bid to go at risk 
  • Weinstein: Yeah, the quote seems wrong I guess. 
  • Trader: given how illiquid it is 
  • Trader: sure do u have a two sided market? 
  • Trader: or what is an appropriate quote? 
  • Weinstein: I guess if you only care at 50 on 2mm then probably 65/ for any size is wrong. 

So we have reliance, for pricing purposes, on information produced by traders who are not really willing to meet their quotes, and whose quotes may differ depending on the size of the investment, and are therefore not well-tailored to depict the hedge fund's specific investment size. The quotes are just that, quotes.  Nothing more.

The concern, then, is that the next softening in the market will also be magnified by our pre-existing market's structural deficiencies: the issues of illiquidity, and our ability to cater appropriately for them in our pricing procedures, could again magnify the uncertainties at play and exacerbate the downturn.

Right now, the price is just not right for illiquid assets, and prices are not consistently applied across firms.  We are ill-advised to think that the prices presented are in any way reliable, given the clumsy (antiquated?) nature with which those prices are derived.

More on this topic soon.
~PF2

Wednesday, March 15, 2017

Can Deregulation and Open Data Solve the Credit Ratings Problem?

This blog is provided by guest contributor Marc Joffe.  The following views are his own, and do not necessarily reflect those of PF2.
~~~   
Credit rating agency scandals, widely blamed for the 2008 financial crisis, now seem to be a distant memory. We have gone several years without another major ratings failure, so casual observers may be forgiven for thinking that the underlying problem has been solved. But as a new Brookings study shows, the defective rating agency market structure that triggered the crisis remains in place. Further reforms would seem unlikely under unified Republican government, but bipartisan support for open financial data may offer a way forward.
Since 2008, the government has taken several steps to address credit rating agency problems. In the waning days of the Obama Administration, the Department of Justice and State Attorneys General settled complaints against Moody’s for $864 million. This followed a larger settlement with S&P and a series of regulatory changes spurred by the 2010 passage of Dodd Frank. That law mandated the removal of credit ratings from regulations, tighter control of SEC-licensed Nationally Recognized Statistical Ratings Organizations (NRSROs) and an SEC study of possible changes to the way investment banks choose rating agencies to rate newly-issued structured finance securities.  
Writing for Brookings, former CBO Director Alice Rivlin and researcher John Soroushian criticize the SEC for failing to execute its Dodd Frank mandate to change the rating agency business model. By not acting, the SEC has left in place a regime under which rating agencies have an incentive to competitively dumb down their standards so that they can sell more ratings to bond issuers. Rivlin and Soroushian recommend that the SEC implement a process under which new structured finance issues are randomly assigned to rating agencies.
Joe Pimbley, a risk analyst who  - like me - used to work at a credit rating agency, goes even further, calling for an outright prohibition of issuer payments for credit ratings. This change would oblige rating agencies to serve investors first, as they did in the years before the transition to the “issuer pays model” around 1970. (Pimbley and I both have consulted for PF2 Securities, which publishes this blog).
Such interventions would seem unlikely under a Republican-led government. If anything, President Trump and Congressional Republicans have expressed the desire to roll back many aspects of Dodd Frank. But, led by Congressman Darrell Issa, Congressional Republicans have shown an interest in more open financial data – and this could be a way forward toward further reform.
To understand the relevance of open data, we must first realize that the credit rating business is not a standalone industry. Instead, the rating agencies are part of a larger industry: the business of credit risk assessment.  This field includes in-house credit analysts at banks, independent credit advisors, and analytics firms, as well as the NRSROs.
By mandating the elimination of credit ratings from federal regulation, Dodd Frank has helped to level the playing field between rating agencies and alternate credit assessment providers. (But, as Rivlin and Soroushian remind us, this process of removing credit ratings from regulations is incomplete.)
Deregulators can go further by pursuing Pimbley’s suggestion of removing credit rating agencies from the list of entities that can receive non-public disclosures from securities issuers, as provided under Regulation FD. Such a reform would allow credit analysts not employed by rating agencies to receive all of the same data at the same time as their agency counterparts.
Indeed, Republicans could completely eliminate the special status of credit rating agencies by scrapping the NRSRO certification entirely, as recommended by NYU’s Lawrence J. White. If ratings are not required by regulation and all credit consultants and analytics firms have equal data access, there would seem to be little benefit to NRSRO status anyway.
But even without regulatory-conferred privileges, rating agencies would still have an advantage over upstart providers of credit analysis. The incumbents’ size allows them to invest in systems and manual procedures to assimilate the large volume of issuer and security data needed to rate and review a large number of debt issues.
Reforms that lower the costs of collecting this data would enable new analytic firms to compete against incumbent rating agencies despite their relatively small size. Representative Issa’s new bill, the Financial Transparency Act of 2017 (HR 1530), would make all financial regulatory data available in machine readable form. Right now, much of this data is only available in PDFs which are costly to process. By instead providing financial filings in the form of structured text, new credit data sets will become more readily available at little or no cost.
One regulator affected by the Act is the Municipal Securities Rulemaking Board (MSRB) which oversees the municipal bond market. Right now, offering materials and continuing disclosures such as annual financial reports are published as PDFs. Anyone hoping to analyze them must either mine the PDFs for relevant data or buy data sets from third parties, usually at high costs and with tight restrictions on redistribution (effectively preventing smaller firms from showing how their opinions are driven by issuer fundamentals). If the MSRB switches to structured text, the cost of analyzing municipal securities would drop, making it easier for municipal analytics startups such as MuniTrend to provide insight across a broad range of instruments.
The MSRB is one of ten regulators affected by the proposed act.  If passed and implemented, the bill would trigger a wave of free and low cost data sets that could help analysts outside of the credit rating agencies keep up with these powerful incumbents. When combined with reforms that remove the special privileges now enjoyed by licensed NRSROs, open financial data could usher in a new era of competition and innovation in the field of credit assessment.