A Gap In Gap Risk

From three basis points to 35 basis points on $10 billion…
sellers began screaming, “Yahtzee!

by James Parascandola, NewOak

On May 26, 2015, the Securities and Exchange Commission (SEC) charged Deutsche Bank with misstating financial reports pertaining to products that were “en vogue” prior to the 2008 financial crisis–namely leveraged super senior (LSS) tranches.

Deutsche Bank agreed to pay a $55 million penalty and neither admitted or denied the SEC’s findings. LSS tranches arise from the residual, senior-most risk via a securitization. Dating back to the late 1990s, investors began to view this principally protected, default-remote synthetic investment as a way to clip a free-money coupon.

Primarily on the menu of insurers, large Yankee banks, reinsurers and monolines, LSS tranches typically carried AAA/Aaa ratings and were initially considered cheap at one to three basis points!

Representing the largest tranche of a securitization by notional, LSS tranches were considered by investors to be principally money good, and protected due to the depth of subordinate tranches that would need to be wiped out in order for the LSS tranche to be principally eroded.

And while this tenet held true apart from mortgage product-based securitizations, what caught many investors off guard was the spread widening that occurred leading up to and through the 2008 financial crisis when sellers of protection on those AAA/Aaa-rated LSS tranches couldn’t be found below 35 basis points. From three basis points to 35 basis points on a $10 billion securitization referencing $100 billion of underlying collateral and protection, sellers began screaming, “Yahtzee!”

That’s where gap risk, or the risk resulting from price changes in the absence of any trading referenced in the SEC’s complaint against Deutsche Bank, came into play. And while attempting to quantify or account for gap risk on bespoke LSS tranches may keep risk managers and regulators awake at night, these trades had a characteristic which challenges the essence of gap risk, leaving regulators with a conundrum: how do you measure gap risk on a bespoke product that has a characteristic of being traded only one time?

Let’s break it down

Selling default protection requires more margin/collateral than buying. When Deutsche Bank purchased LSS protection, they required their counterparts to post approximately 9% collateral as margin; not an unseemly or obviously deficient figure considering it was common market practice to require sellers of protection to post anywhere from 15%-0% for selling protection on high-yield and cross-over rated reference entities.

After all, participants were not selling protection directly on single B-rated reference entities trading at 1500 basis points, they were selling protection on Aaa-rated tranches trading at three basis points. According to the SEC’s findings, Deutsche Bank tripped itself up in their reporting and modeling methodologies, which were perhaps just as deficient as the entire market’s margin protocols at the time.

Margin market protocols, which were suspect on the most liquid and observable single name and index credit derivatives, let alone on a product that had virtually no secondary market, were even more a function of guesswork for bespoke LSS tranches where transparent and observable pricing simply didn’t exist.

A 500 point swoon

A 500-point morning swoon in the DOW in 2008 could have sent LSS tranches wider by five to 10 basis points as CDX.IG spreads approached their all-time wides of nearly 300 basis points. All good things for Deutsche Bank and all buyers of protection unless, as the SEC alleged, you fail to mind your gap risk – which in this case is the difference between the in-the-money amount of Deutsche Bank’s credit protection and the value of the collateral pledged by the protection sellers.

For instance, given the leverage on the deal referenced above, a one basis point widening in underlying should equate to an approximate 10 basis point move wider in securitized product, depending on a variety of factors, including tranche attachment/detachment point, underlying defaults and overall market liquidity.

The notional amount of Deutsche Bank’s protection was $98 billion; the amount of collateral pledged by sellers was reported to be $880 million, an amount equal to roughly a 1.8 basis point move wider in LSS tranche pricing assuming a 5-year average duration. If we conservatively assume LSS tranche spreads widened to 35 basis points, this should have called for an additional $16.1 billion in margin collateral to account for the mark-to-market move in Deutsche Bank’s favor.

Deutsche Bank estimated their gap risk to be between $1.5 billion to $3.5 billion for internal purposes and adjusted the net present value of their protection to an amount equal to that of the collateral pledged; effectively reducing gap risk to zero for reporting purposes, according to the SEC’s complaint.

When considering this particular case, it’s important to note that market conditions around the most liquid credit instruments deteriorated markedly throughout 2008. And ascertaining precise valuations for illiquid, leveraged instruments amidst the most violent price swings in the history of the credit derivatives market was guesswork at best.

So while one could challenge the magnitude of underreported gap risk, the SEC’s findings establish that Deutsche Bank was deficient in their reporting and fined them $55 million, a fractional amount compared to the more than $1 billion in fines levied by U.S. and European regulators against JPMorgan for that firm’s controls around credit derivative index trading within their Chief Investment Office division, where losses reached $6.2 billion.

Also Considered:

Tibble vs. Edison: Ripple effect for plan sponsors, fund managers

by Timothy Bernstein, NewOak 

Perhaps the most notable thing about the Supreme Court’s unanimous decision in Tibble v. Edison International was its inevitability.

By the time the court ruled in favor of participants in Edison International’s 401(k) plan, the argument had essentially become a debate of semantics revolving around the scope of one word: “prudence.” What the plaintiffs argued before the court was that managers of Edison’s retirement fund must actively monitor and review the prudence—or imprudence—of investment decisions they made and adjust if necessary.

Edison, in response, argued that the statute of limitations on any alleged fiduciary breach had been exhausted by 2007, when the company’s employees filed the suit (the claim centers around three retail-class mutual funds added to the retirement portfolio in 1999 when there were allegedly cheaper “wholesale” funds available for purchase instead).

Edison felt that its fund managers need only reevaluate strategies when they made significant changes to the fund, while Edison’s employees felt that more consistent monitoring was appropriate or, in this case, prudent. Prudence, after all, is central to a manager’s fiduciary responsibilities under the Employee Retirement Income Security Act (ERISA). ERISA guidelines on the Department of Labor website describe the term succinctly as “the process for making fiduciary decisions.”

A 'Continuing Duty'

Justice Stephen Breyer’s opinion went one step further, writing that ERISA dictated that a fund’s trustee has “a continuing [emphasis mine] duty to monitor trust investments and remove imprudent ones”[1] and that “this continuing duty exists separate and apart from the trustee’s duty to exercise prudence in selecting investments at the outset.”

The High Court, which has established a strong precedent of interpreting a trustee’s ERISA-mandated obligations liberally, needed little convincing to reinforce the need for continuous monitoring. By the time the case went to the Supreme Court, even the defendants had difficulty contesting the central issue.

As Ronald Mann wrote on SCOTUSblog.com, “The result has not really been in doubt since Edison filed its brief earlier this spring agreeing that the duty of prudence requires periodic monitoring.”[2] In the bigger picture, the question remains as to what the long-term significance of this ruling will be for the investment management industry.

A significant number of advisors to plans like Edison’s 401(k) are not officially designated 3(38) ERISA fiduciaries to those plans; if they were, they would be legally liable for the continuous choosing and monitoring of the investments made in those plans.

The Court’s decision in Tibble indicates just how far that liability might stretch (no official penalty for Edison has been announced yet, but the plan’s advisors have been found liable for investments they approved in the previous century).

The Department of Labor has put a proposal on the table that would require every advisor to a retirement plan to be an official ERISA fiduciary, but whether or not those rule changes go into effect, the Supreme Court’s unanimous decision may send enough of a message to retirement plan sponsors (in this case, that was Edison itself) to gravitate away from non-fiduciary advisors and towards ones legally required to respect the “continuous” nature of monitoring.

Even if this shift does not occur, the decision might at least add momentum to the ongoing campaign by plan sponsors and participants for increased transparency in fund management fees. The central difference between the retail-class funds that Edison’s 401(k) managers purchased in 1999 and the cheaper wholesale funds is that the former carried higher management expenses. It took eight years and the Supreme Court to declare this conduct unprofessional in the context of a monitor’s obligations. If enough people take away the salient message from Tibble v. Edison and a 9-0 decision, there is a chance this lesson will not need to be taught again.