Secretive practices and institutions are common in contemporary finance. For those who’ve ceased the search for long-term value creation, temporary information advantage is key. Even commonplace practices can be reinterpreted as havens of hiddenness. My colleague Michael Simkovic’s article “Secret Liens and the Financial Crisis of 2008” exposes the role of derivatives and securitization as secretive borrowing strategies, designed to keep the naive or trusting from discovering the fragility of the institutions they loan funds to. His work has been presented to the World Bank Task Force on the Bankruptcy Treatment of Financial Contracts, and is relevant to both private and sovereign debt risks.
Simkovic argues that 80 years of erosion of classic commercial law doctrine ensured that “complex and opaque financial products received the highest priority in bankruptcy.” Products like swaps and over-the-counter derivatives were not adequately disclosed (either by banks in their consolidated financial statements or by their counterparties in publicly accessible transaction registries). By concealing those debts, these already overleveraged financial institutions were able to attract ever more credit and investment, at better rates than those who reported their overall financial health more accurately. (All other things being equal, it’s safer to lend to an entity that owes 10 billion rather than 100 billion dollars.) The genius of Simkovic’s article is to show how “fundamental causes of the financial crisis are relatively old and simple,” even as an alphabet soup of instrument acronyms (CDO, CDS, MBS, ad nauseam) and government programs (TARP, TALF, PPIP, et al.) makes our time seem unique.
As Simkovic explains:
Losses act as a spark; widespread leverage is the powder keg. Leverage can be “regulated” privately by creditors or regulated by government, but only if the extent of leverage is known [emphasis added]. Hidden leverage is a perennial problem because debtors rationally wish to borrow at the lowest price possible. Debtors can borrow at more attractive rates by hiding their existing debts and creating an exaggerated appearance of creditworthiness. [emphasis added]
Debtors who wish to hide their debts can exploit competition between potential creditors to gain active cooperation from some creditors. These cooperative creditors will work with debtors to hide loans either through simple non-disclosure or through complex structures. Debtors may compensate these cooperative creditors for their assistance with higher fees, a deeper business relationship with the creditors, or liens on the debtors’ property. The result of this subterfuge is lower financing costs for the debtor and lower profits—or steeper losses—for unsophisticated unsecured creditors.
Note that Simkovic’s work is more incrementalist than that of Stephen Lubben (another colleague of mine) or Mark Roe, who question the wisdom of safe harbors for derivatives in bankruptcy. Whatever you feel about that position—and however you feel about the relative advantages of regulation or market forces in deterring systemic risk—Simkovic’s work points to a fundamental problem that all sides in these debates must grapple with. Neither market forces nor regulators can deter systemic risk if there’s not fair warning that an interaction between secrecy and priority in bankruptcy can suddenly create disastrous runs on financial institutions. It’s one thing to create priorities (or supersecured creditors) that everyone knows about. It’s quite another to allow sophisticated debtors to promise the moon and stars to entities that have no idea what rival claimants are going to demand.
Pushing for Priority
For “financial innovators” in the years leading up to 2008, the game was straightforward: sophisticated lenders wanted to obtain first priority in bankruptcy (and/or the right to make collateral calls quickly), while borrowers wanted to hide how much they’d borrowed (and how encumbered their assets were). Like the fraudulent subprime broker who added a few zero’s to the end of his client’s W-2 form, leading bankers exaggerated the well-being of the desks and divisions they fronted for by obscuring certain obligations on their books.
How did they do it? Simkovic explains how the bankruptcy code now favors “securitized” over “secured” debt. The code forces secured creditors to try to keep their bankrupt debtor afloat during reorganization. For example, an “automatic stay prevents a secured creditor from seizing and liquidating the underlying collateral to recoup its investment.” A securitization can stave off such obligations by “distancing” certain obligations from bankruptcy. Simkovic explains these steps:
In an asset securitization, the debtor (or “Originator,” the term typically used in documentation) transfers financial assets such as credit card receivables or mortgage receivables to a special purpose entity, or SPE, typically a wholly-owned subsidiary of the debtor. The SPE (or another transferee) issues debt to investors. Investors pay the SPE which then pays the debtor.
For the securitization to isolate the underlying assets from the debtor’s bankruptcy, the transfer of assets from the debtor to the SPE must qualify as a “true sale.” Most securitizations do not economically resemble “true sales” because the debtor retains the risk of default or non-performance of the underlying assets. The debtor retains risk because the debtor owns the equity (or “first loss tranche”) in the SPE, and because the debtor may be required to repurchase assets from the SPE if losses reach a level exceeding . . . [a] pre-set trigger.
However, just as clever legislators let AIGFP characterize its disastrous CDS business as “protection-selling” (rather than insurance), Delaware’s Asset-Backed Securities Facilitation Act let securitizers safely call the SPE fancy footwork a “true sale” to avoid the responsibilities associated with secured debt. The debtor’s obligation to its SPE’s is kept “off balance sheet,” hidden from many creditors. Limited disclosure of asset securitizations (and their terms) means that “even professionals can underestimate the extent of debtors’ exposure to losses from securitized assets.”
Rating agencies have given very high ratings to securitized debt, reasoning that the originating “‘companies retain the subordinated interest in the transaction known as the equity tranche or ‘first-loss’ piece.'” But they ignored the underlying economic motivations behind the transaction: those who brokered the deals would walk away with huge fees regardless of how well it did overall. Aside from sacrificial wolf Fabrice Tourre, virtually everyone involved in the securitization machine has done fine financially. And as the WSJ noted in 2008, the SEC failed to require rating agencies to “disclose to the public all underlying information about any debt they are rating.”
Financial Innovation as Epiphenomen of Legislation: The Case of Credit Default Swaps and BAPCPA
Simkovic also highlights how another aspect of the CDO doomsday machine thrived on secrecy. Just as a firm could both stand behind a securitization (to assure buyers of the securitized assets) and not stand behind it (for accounting purposes, so it looked like it had less exposure than it actually did), so too could the buyers of securitized assets have their cake and eat it too. The securitized asset promised a steady income stream, and a transaction called a “credit default swap” allowed its beneficiary to offload the risk that the income would not materialize onto another entity, in exchange for steady payments of its own. So even when outside observers might bridle at the amount of leverage an entity took on to buy securitized assets (often from another entity that was heavily leveraged to create and support the same assets), the buyer could outwardly appear to be placing one bet with its publicly disclosed balance sheet, while secretly hedging its bets by buying a credit default swap from a well-capitalized firm that promised to pay in case the SPE and originator could not. Its purchase of the assets, a “vote of confidence” in public, might even be swamped by skepticism about the viability of the assets, if the CDS paid off far more than the expected value of the CDO it insured.
By September, 2008, AIG had sold $440 billion of CDS protection. It had no way of paying out anywhere near that amount, and had not reinsured itself, or offloaded some of the risk onto someone with deeper pockets. Where were the regulators? Stupefyingly uninformed, as the FCIC Report shows:
The Office of Thrift Supervision has acknowledged failures in its oversight of AIG. . . . [S]upervisors failed to recognize the extent of liquidity risk of the Financial Products subsidiary’s credit default swap portfolio. John Reich, a former OTS director, told the FCIC that as late as September 2008, he had “no clue—no idea—what [AIG’s] CDS liability was.” [emphasis added].
According to Mike Finn, the director for the OTS’s northeast region, the OTS’s authority to regulate holding companies was intended to ensure the safety and soundness of the FDIC-insured subsidiary of AIG and not to focus on the potential impact on AIG of an uninsured subsidiary like AIG Financial Products. Finn ignored the OTS’s responsibilities under the European Union’s Financial Conglomerates Directive (FCD)—responsibilities the OTS had actively sought. (350) [emphasis added]
Throughout late 2007 and 2008, the company’s accountant tried repeatedly to discover the true amount of risk involved in its transactions. As Simkovic shows, this was exceedingly difficult to do:
Credit default swaps, like most OTC derivatives, are an ideal vehicle for hidden leverage and secret liens because of their inherent complexity, limited disclosure, and superior treatment in bankruptcy. . . .Unlike exchange traded derivatives, which are standardized, simplified, and priced by the market through frequent trading, OTC derivatives are custom, bilaterally negotiated, relatively illiquid contracts and therefore difficult to price. The value of the derivative depends on three things: (i) the value of the underlying asset; (ii) the contractually negotiated formula that determines the counterparties’ obligations to each other based on that value; and (iii) the creditworthiness of the counterparty to the derivative, which determines the likelihood that the obligation will actually be paid. . . .
In the case of credit default swaps written on CDO tranches held by financial institutions: (i) the value of the underlying assets is difficult to determine because of the mathematically complex structuring that governs loss allocation among tranches and because of limited information about the credit quality of the underlying loans; (ii) the extent of counterparties’ obligations to each other is difficult to determine because of the subjective nature of determining when a “credit event” has occurred and the risk that disagreement will result in litigation; and (iii) the creditworthiness of counterparties is difficult to determine because they too have extensive and hard-to-measure exposures to derivatives such as credit default swaps.
Simkovic discusses how it may be impossible, even in principle, for large players to figure out the true extent of their exposure. CDS counterparties thought they were safe once they bought protection from AIG, and didn’t realize that AIG might go under. The banks didn’t accurately gauge the risk posed by AIG.
Simkovic’s position has been amply confirmed by other critics’ work. Consider, for instance, the cutting analysis from Eric Banks’s prescient The Failure of Wall Street (2004), which describes “financial controllers and auditors who don’t understand the nature of the business they are meant to be ‘independently monitoring,'” and trading desks which have little sense of “what kind of credit risks they are exposed to.” Combine these internal weaknesses with the regulatory arbitrage that allowed institutions to seek the most pliant “watchdogs,” and disaster was inevitable.
Of course, there were some forward-thinking participants in the financial markets who saw the risks, and ran away from them. As one report noted about Warren Buffett:
When Berkshire bought General Re in 1998, the reinsurance company had 23,000 derivative contracts. “I could have hired 15 of the smartest people, you know, math majors, Ph.D.’s. I could have given them carte blanche to devise any reporting system that would enable me to get my mind around what exposure that I had, and it wouldn’t have worked,” [Buffett] said to [a] government panel. “Can you imagine 23,000 contracts with 900 institutions all over the world with probably 200 of them names I can’t pronounce?” Berkshire decided to unwind the derivative deals, incurring some $400 million in losses.”
As one commentator observed, this type of complexity leads to a number of other problems:
Derivatives, because they are so hard to value, make it easier for traders and chief executives to inflate earnings. They exacerbate problems if a company, for unrelated reasons, suffers a credit downgrade that requires it to post collateral with counterparties – “a spiral that can lead to a corporate meltdown,” [Buffett] wrote. They create a “daisy chain” of risk as the troubles of one company infect another.
That “daisy chain” of risk echoes the diagnoses of Yale economist John Geanakoplos (whose work has indicated the instability caused by high leverage and “tight chains of financial interdependence”), Rick Bookstaber (who chronicles the instability inherent in “tightly coupled” systems), and Lawrence Baxter (who brings attention to recent network science on the “cascading failure” that is “common to many complex systems”). An interdependent system as complex as the OTC derivatives market can crumble at any moment. If key participants are too highly leveraged and one or more of them experience a shock, disaster is inevitable. The problem is not a “black swan;” it’s the black box dealmaking that make it impossible for markets or regulators to grasp how leveraged and fragile institutions really are.
Secrecy vs. Resilience
So how resilient should these systems be? That is a matter due much more political and regulatory attention than it is currently getting. Regardless of one’s views on leverage, one thing is clear: the types of opacity described in Simkovic’s article prevent us from getting any handle on the scope and severity of the problem. As he notes,
[E]ven basic information about OTC derivatives transactions can be extremely hard to come by. Market participants themselves are often unaware of the extent of their net exposures or the identity of counterparties to their transactions. Mandatory disclosures to third parties are even more limited, and the industry group, the International Swaps and Derivatives Association, has resisted voluntary disclosure.
Why the lack of clarity? There’s a fundamental contradiction in finance. Financial managers need to compete by keeping what they know secret so they can place big bets at “wrong” prices and make money when the prices eventually correct. At the same time, the policy justification for financial markets is that markets get the pricing right. As Simkovic has argued, “Allowing greater secrecy is essentially a decision to allow financiers greater private profits and to reduce the public benefit from quick ‘price discovery’ by markets.”
Accounting rules compound the difficulties, allowing certain deals to be quarantined from the rest of the bank’s financial status. For example, since a “simple fixed/ floating interest-rate swap contract . . . has zero value at the start,” it “is considered neither an asset nor a liability, but is an ‘off-balance-sheet’ item.” Lehman Brothers had $738bn in derivative contracts labeled as “off balance sheet arrangements” in its 2007 accounts, it’s now hard to accept uncritically its estimate of their ultimate “netted” value at the time. Carol Loomis did a post-mortem on the situation:
Lehman had a derivatives book of only $730 billion as it neared bankruptcy. Even so, when Lehman’s U.S. entities filed for Chapter 11 in September, this not-so-big figure translated into about 900,000 derivatives contracts. The great bulk of them have been “terminated” by derivatives counterparties which under industry protocols had the right to immediately “net” their accounts with Lehman in the event it declared bankruptcy. A handful – the last reported number was 18,000 – are still open.
Each of these contracts has a “fair value” – an amount that one side owes the other. Lehman, in fact, has a lot of open contracts that have been going its way. In a droll sign of how derivatives have come to be viewed as indispensable, Lehman has received permission from the court to buy them to hedge some of its open contracts, so that it can lock in the profits it has made since filing for bankruptcy.
Move now to the accounting problem. While sometimes the fair value of a derivative can be precisely determined, at other times it must be derived from murky markets and models that leave considerable room for interpretation. That gives the holders of derivatives a lot of bookkeeping discretion, which is troubling because changes in fair value flow through earnings — every day, every quarter, every year — and alter the carrying amounts of receivables and payables on the balance sheet.
The subjectivity involved in derivatives accounting also means that the counterparties in a contract may come up with very different values for it. Indeed, you will be forgiven if you immediately suspect that each party to a derivatives contract could simultaneously claim a gain on it — which should be a mathematical impossibility. In fact, we have a weird tale, gleaned from court documents, supporting that suspicion. It involves Lehman, Bank of America, and J.P. Morgan, and suggests how far some of those “terminated” contracts are from being truly settled.
That last point—that both parties could “simultaneously claim a gain” on what had to be zero-sum arrangements—is critical to understanding the risks posed by black box finance. It explains why deal complexity is often pursued for its own sake, and not for a genuine economic or investment purpose. Webs of debt become a smokescreen for what is really going on: institutions are rendered more and more vulnerable (both individually and collectively) so that privileged parties within them can reap enormous incomes from fees and bonuses. Formalities didn’t matter: as Stephen Lubben notes, “many credit default swaps were assigned to new protection buyers without the prior consent of the seller,” even though the ISDA Master Agreement governing such deals forbids this. Murky accounting kept Chuck Prince’s famous “music” playing ever longer, as a mountain of leverage and misallocated capital accumulated.
It Gets Worse: BAPCA’s Ugly Legacy
According to Simkovic, 2005 amendments to the Bankruptcy Code under the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (“BAPCPA”) have exacerbated the problem. This law assured that “derivatives counterparties effectively bear no risk of loss to the extent that the debtor posts collateral to cover its obligations.” Simkovic describes the rush of creditors to take advantage of its provisions, a stampede that made it “harder for those creditors to communicate with one another and monitor debtors’ leverage.” AIG was the poster child for overleveraged indebtedness, selling $440 billion in “protection” on CDOs. According to Lynn Turner, “In one year, the disclosures from the company had gone from not losing a dollar to over $26 billion in valuation losses and counter parties that to this day have not been disclosed demanding over $16 billion in collateral.” Since the Treasury Department believed that “the global economy was on the brink of collapse” when the the magnitude of the problem became clear, the government stepped in to bail out AIG (and, thus, its counterparties).
(Turner, former chief SEC accountant, describes how opaque AIG’s procedures were. The recent book Fatal Risk gives the “tick tock” details, chronicling the deepening unease of AIG’s auditors as the mark-to-fantasy approach of its subsidiary AIGFP became clear.)
From 2000-2008, AIG made $66 billion in profit, but in 2008, it had a $99.3 billion loss. The employees and execs who benefited in the boom years have kept nearly all their cashed out compensation. By January, 2011, according to the FCIC report, AIG had cost US taxpayers $152 billion (FCIC Report, 350). It’s an incredible sum for a process whose only social contribution, so far as I can see, was a marginal (and likely temporary) bump upwards in the rate of homeownership.
Simkovic describes financial sector creditors consumed by a desire for positional advantage, ignoring the slow erosion of the viability of the system as a whole. It’s an “I’ll be gone, you’ll be gone” culture, where accountability has largely vanished. If we really want to understand the recent investor rush to gold and commodities, it’s better to look beyond the central banks’ “printing money” and to think hard about why they’ve had to do so. Do you really want to park much money—be it in stocks, bonds, or some other instrument—at institutions staffed and run by people who bear virtually no pain in case of their collapse?
Picking Up Pennies in Front of a Bulldozer
Asymmetric compensation schemes are common in finance: managers enjoy substantial upside (perhaps even a life of leisure) if things go well; there is very limited downside if things go badly. Simkovic analyzes how opacity, corporate law, and bankruptcy code provisions help preserve this lack of accountability at the core of finance. The usual Wall Street metaphor for behavior like AIG’s is “picking up pennies in front of a bulldozer.” But the agency problems (and amounts involved) flesh out the metaphor: financial institutions are like wheezing couch potatoes, suddenly running to pick up suitcases of cash in front of a bulldozer of risk, and delivering most of the proceeds to high-flying managers lounging on settees by the side of the road. When the financial institution finally ends up getting a hand or arm caught under the bulldozer, it faces the horror of an impromptu amputation or annihilation. The managers will surely rue the untimely death or disability of their “runner,” but they’ll walk away with cash they’ve “earned,” and almost certainly find some other institutional form to renew the game another day. Even if some don’t, they may have made enough while the getting was good to fund an early retirement to the Caribbean.
At a recent event on the LSE report on the Future of Finance, panelists and audience members suggested that rent-seeking, as well as tax, accounting, legal, and agency distortions, are driving finance sector transactions more than the real economic substance of deals. Simkovic helps us understand the full extent of the problem. The bankruptcy code is becoming the tail that is wagging the dog of investment decisions. He argues that we need to “revive recordation,” as Hernando de Soto and other luminaries have urged. Perhaps real financial reform will ultimately depend on more radical approaches. But I see no way of significantly improving the situation without regulators taking seriously the problems Simkovic has described. In my next post on the issue, I will look at Michael Greenberger’s relatively optimistic take, and Arthur Wilmarth’s pessimistic view, on whether aspects of Dodd-Frank can address Simkovic’s concerns. Early developments have not been promising.