Much has been made of the value of liquidity in the marketplace and its role in preventing the kind of melt-downs that, hitherto, have caused anything from fifteen-standard deviation volatility to recessions to the election of William Jefferson Clinton (depending on who you ask). Dot-comesque analogies ("it's different this time," "the market dynamic is changed," "new paradigms," "the old rules don't apply") are, however, the first symptom of overreaching by the more optimistic among us and the one that causes astute Going Private readers to start looking for clever and leverageable shorting opportunities. The simplistic view is easy enough to articulate that you can use a single operator sign. Liquidity = Good. Rationally skeptical Going Private readers (a redundant description) recoil at such simplistic formulae. Could there not, they might ask, be a dark side to liquidity?
I think, as we overlay the question of liquidity on top of the current credit landscape (and for selfish reasons I have a particular interest in the corporate debt/commercial paper side), it might be helpful to trace the impact of liquidity and other structural issues on debt instruments in each step down (and up) stream from demand to supply to securitization.
In this connection, a series of articles have come to my attention that shed some light. One of the things that frustrates me about "financial journalism" is that these kinds of structural analysis are few and far between. They require deeper and more fundamental thought, and a certain disciplined approach that, even if a member of the press mustered the motivation, is possessed of a vanishingly small audience. (c.f. Vipal Monga over at The Deal who should be on every Going Private reader's "must read" list). Accordingly, such inquiries often require the assimilation of a number of disparate sources into a larger whole. In this case, the result will initially resemble a large linkfest, a style of posting I generally have an aversion to (c.f. Abnormal Returns). Patience, dear readers. Patience.
I trust that a discussion outlining the basic demand and the structural incentives for the first layer of debt (be it a home mortgage or corporate debt to finance an LBO) need not occupy much space in these pages. Going Private readers will be intimately familiar with the effect of cheap cash and debt on these demand curves. They should also recognize that liquidity in such quantities necessarily has the effect of reducing the returns (quality) required of a transaction in order to trigger a decision to invest. Lower costs of capital result in pulling the trigger on deals with lower returns.
A flood of liquidity into such deals removes the "low hanging fruit" and reduces the universe of profitable opportunities. Further, the delay in return reporting (reports of LBO returns significantly lag the last investment cycle due to the "J curve" effect of LBO investment) may cause liquidity in LBO spaces (or elsewhere) to surge just when opportunities are becoming scarce. The result may be an imbalance between liquidity (the willing supply of cheap debt, pressed into service by the demand for the debt returns by issuers) and the actual availability of quality deals. Going Private readers will recognize that the pressure to "place debt" on the part of LBO sponsors will drive more and more creative (foolish) investment theories as firms struggle to deploy cheap credit into dwindling opportunity pools.
What might bear examining more closely is the next level, the incentives for banks to shift that debt off their own books. For this. The Wall Street Journal's David Wessel points out:
One culprit, however, has gone unnoticed: A sweeping change in international rules governing the capital banks must hold. By requiring banks to boost the capital held in reserve against the loans carried on their books, the rules encouraged banks to get rid of those loans by turning them into securities to be sold to investors. Banks took the hint.
Via the always yummy Abnormal Returns last week, we were pointed to this piece from The Big Picture, outlining a reader's take on the irrational (or insidiously rational) behavior by CDOs. Specifically:
I was talking to CDO managers in mid-'05 that were saying how rich sub-prime MBS was and how wrong everyone was for buying that stuff at the spreads they were. To a man, they all agreed they were paying too much for the risk, they all believed that [home price appreciation] was going negative soon. But, sadly, they had to buy the stuff because they needed to accumulate collateral for their CDO issuance. Fuck, we all knew we were overpaying, even back in 2005. We knew it was essentially a bet that home price appreciation was going to continue at levels that couldn't be sustained. No way that could keep going on.
So what was the prime motivating factor?
The answer is quite simple: DEAL FEES. I gotta keep buying collateral, in order to keep issuing these transactions as a CDO manager. Its my job: I gotta keep accumulating collateral, and I gotta issue the liability against that collateral.
This is an interesting "down-stream" bit of color from a moral hazard issue in loan origination that I reflected on back in March. Specifically:
The Going Private reader, being well versed in such matters, will recognize this as a "moral hazard" problem. As a Vice President, why not write the big loan today that will likely blow up in 48 months? Bonus time is only 9 months away, after all. Absent a claw-back provision for such bonuses, why worry? And, taking it a step further, if you are the managing director of a bank filled with Vice Presidents writing loans that may well blow up in 48 months, why not sell them to a CLO? (A nicely done, albeit slightly old, primer on CLOs of exactly the sort that will intrigue the Going Private reader and written by Andreas Jobst can be found here and is recommended).
I will bore long-time Going Private readers via redundancy by pointing out that the essence of moral hazard problems is a disconnect between information and compensation or incentive structures, of the kind that will cause Going Private readers to cringe when non-risked based insurance premiums are proposed. (For this reason, Going Private readers have a particular distaste for FDIC-like programs, which were not effectively risk based- i.e. were willfully on the wrong side of asymmetric information imbalances- until recently). A particularly relevant problem for our purposes is directly related to moral hazard: the "principal-agent" problem, also related to information asymmetry. In this case, the agent is acting on behalf of the principal, but the principal lacks information about the agent's intentions or interests. In the absence of careful monitoring or a mechanism to align the agent's incentives (or disincentives) with the principal's interests, trouble brews.
The Alea blog then delivers two interesting, and seemingly unrelated, articles pointing to academic papers that round off the discussion. First, they tackle the question "Why Are Securitizations Issues Tranched?" resulting in this, interesting conclusion:
Finally, we investigate the effect of tranching on the pricing of issues at launch. We find evidence that tranching might be successful in remedying problems of market segmentation. For those issues where our model predicts a higher optimal number of tranches, we find that additional uniquely-rated tranches are associated with higher prices for the issue as a whole. This suggests that structuring is allowing issuers to exploit market factors – such as greater investor sophistication and heterogeneous screening skills related to asymmetric information – to their advantage via tranching.
Not to be left behind in the liquidity discussion, they then provide us with "Asset Pricing in Markets with Illiquid Assets," presenting a conclusion that both will and won't surprise Going Private readers:
Illiquidity has a dramatic effect on optimal portfolio decisions. Agents abandon diversification as a strategy and choose highly polarized portfolios instead. The value of liquidity can represent a large portion of the equilibrium price of an asset. We present examples in which a liquid asset can be worth up to 25 percent more than an illiquid asset even though both have identical cash flow dynamics. We also show that the expected return and volatility of an asset can change significantly as the asset becomes relatively more liquid.
Every bond portfolio is marked to model. Every mutual fund. Every UIT. Every brokerage account. Every portfolio manager. Every pension. Every hedge fund. Every bond portfolio in the whole damn world is marked to model.
The article goes on to say:
- There are 152,732 different US dollar denominated corporate bonds outstanding as of today, according to Bloomberg.
- Between Fannie Mae, Freddie Mac, Federal Farm Credit Bank, and Federal Home Loan Bank, there are 21,989 different issues outstanding.
- There are 1,269,428 different municipal bond issues outstanding.
How many of those trade on a given day? A very small percentage. According to TRACE, which is a system for tracking corporate bond trading, about 4,300 TRACE eligible corporate issues trade each day. That's less than 3% of the total corporate universe. I note that trading activity in non-TRACE eligible bonds is not publicly reported. According to the MSRB, about 14,900 municipal bonds trade each day, or just over 1% of the universe.
Completing the Abnormal Returns link trifecta, the same, and still yummy, Abnormal Returns piece suggests we might find a piece by TheStreet.com on the annoying and more recent tendency of gold to be correlated to stocks, and therefore to also be more and more useless as a diversification tool or a safety net. Says TheStreet:
After all, the more people there are who believe gold will rise when stocks fall, the more the price of gold is likely to be bid up ahead of a selloff in stocks.
think this attributes a bit more predictive skill to the general market
than it historically exhibits, (perhaps I err in using TheStreet as a
reference given the oft poor analysis) but the observational data holds
(whatever the analysis in front of it).
What does this mean for the structure of the credit markets? Extremely low credit spreads make a "debt-equity" arbitrage opportunity compelling. Using cheap cash / debt to finance the purchase of equity (particularly the highly liquid equity in publicly traded firms) should be a no-brainer given the spreads. There is also the effect of differing attitudes towards debt in the public and private markets. I discussed this subject back in June. Add to this an illiquidity arbitrage opportunity and demand for LBO debt (indeed any debt) soars. What, you might ask, is illiquidity arbitrage? For this we turn back to Asset Pricing in Markets with Illiquid Assets:
Finally, these results illustrate that differences in patience across investors can have major asset-pricing effects. Thus, heterogeneity in patience may provide an important (but underresearched) channel for understanding financial markets and resolving asset-pricing puzzles. Furthermore, differences in patience may map well into the familiar notions of short-horizon financial institutions and longer-horizon investors often found in the literature.
We can also see from this paper that:
As might be expected, illiquidity generally has the effect of making the liquid asset more valuable, and the illiquid asset less valuable, relative to what prices would be in the fully-liquid case. Surprisingly, however, the opposite can be true. This situation occurs primarily when the impatient agent’s incentive to hold the liquid asset conflicts with his incentive to hold the riskier asset.
If it is true that the public markets are short-term thinkers, it is also true that they seem happy to be such. Liquidity, it seems, may be overpriced. Patient agents are taking advantage. The impatient agent's incentive to hold a liquid asset may conflict with an incentive to hold the riskier asset, no?
And so, seeing a demand to fill, banks issue debt. Quite a lot. But, dear friends, this presents them with David Wessel's quandary. They must hold larger reserves as they take debt on their books. Fortunately, they have a mechanism to shift the liabilities off their books effectively. No, no, not special purpose entities, that's so last year. Securitization.
Moreover, avoiding reserve requirements, while one piece of the puzzle, is not the only one. Moving what could be termed "regulatory overcharged" assets (those assets that count, for instance, towards capital reserve requirements, or have other impacts on the bank's regulatory risk profile- as distinguished from the market's view of the riskiness of the assets) into tradeable securities permits institutions to arbitrage the difference between regulatory risk measures (which tend to be more conservative) with market risk measures of the same securities. Consider also that to the extent banks can reduce the debt on their books, they can reduce their provisions for bad debt and thereby reduce quite directly their cost of capital.
Collateralized Debt Obligations are one of several "asset backed securities" that, as the Going Private reader is surely aware, have seen tremendous growth in the last many years. In essence, a separate entity acquires various debt pieces, bundles them and then issues tradeable securities in various tranches to investors. Astute Going Private readers will recognize this passage from Jobst:
The issued securities are structured in so-called senioritized credit tranches.... The tranching can be done by means of various structural provisions governing the participation of investors in the proceeds and losses stemming from the collateral. Subparticipation is one of the most convenient vehicles for attaching different levels of seniority to categories of issued securities, so that the losses are allocated to the lowest subordinated tranches before the mezzanine and the senior tranches are considered. This process of filling up the tranches with periodic losses bottom-up results in a cascading effect, which conversely applies in the distribution of payments from collateral by the issuer. Both interest and losses are allotted according to investor seniority. Thus, the prioritization of claims and losses from the reference portfolio guarantee that senior tranches carry a high investment-grade rating (triple-A or double-A rating), provided sufficient volume of junior tranches have been issued to shield more senior tranches from credit losses.
To guard against the potential for ex ante moral hazard (i.e. that the CDO issuer would attempt to avoid monitoring costs, etc.) CDO issuers generally hold the riskiest tranch, the equity layer.
A critical part of understanding CDO structures, and the incentives that motivate the selection of assets in the reference portfolio that defines the assets base of the collateral, is the need to blend debt instruments with varied risk/return profiles. Going Private readers will recognize in this the advantage of blending low-risk assets with risky assets to develop a portfolio that sits near the most advantageous part of a risk/return frontier. The importance here? The acquisition of risky debt instruments is essential to CDO portfolio and therefore CDO issuance. AAA rated debt must be mixed in with riskier instruments. The larger the demand for CDO instruments in general the larger the demand for the riskier (but higher return) debt- irrespective of the quality of the debt. A large amount of liquidity pouring into the CDO space can "artificially" press up the prices of this debt as portfolio models force CDO managers to buy more and more collateral at this level. As the prices of the lower rated debt rise due to the increase in demand, CDO managers have to bring aboard even riskier debt to continue to enjoy the returns demanded by their investors. The net result? A slow but certain increase in risk of the collateral underlying CDO instruments.
Turning back upstream for a moment, demand for risky debt necessarily places incentives on originators to create more of it. Hence our Big Picture CDO insider's comment:
To a man, they all agreed they were paying too much for the risk, they all believed that [home price appreciation] was going negative soon. But, sadly, they had to buy the stuff because they needed to accumulate collateral for their CDO issuance.
And let us delve deeper into the internal structure of CDO firms. It seems clear that with this kind of liquidity, incentives for CDO firms quickly become to transform themselves into "placement agents." They are more in the business of minting CDO instruments to collect fees rather than acting as investment managers. Compensation schemes reinforce this basic goal (placement). Bonuses are typically tied to overall fees or the ability to place/issue more instruments (regardless of quality). As with many incentive structures in more "liquid" investment management roles (hedge funds, CDO managers, etc.) these are poorly aligned with the long-term performance of the portfolios. The incentive for the CDO manager is divorced from the interests of the CDO's investors.
Attach to this the reversal of the usual preference for liquidity (in the face of the need to hold the risky asset for portfolio purposes) and you have an artificially inflated market for illiquid assets. This brings full-circle the observation in the Illiquid Asset paper, that liquidity, in excess, can create a flight to what are typically illiquid assets and that the price (value) of assets can be divorced entirely from their cash-flow properties.
There is a third order effect here as well. The prices of these illiquid assets are inflated, they look more liquid than they are and they have the effect of boosting the value of collateral portfolios that "mark-to-market" as the traded prices (or their acquisition costs) are integrated into the books of CDO portfolios. And when the liquidity of these instruments dries up? Their "value" on the books is frozen at the high point by the lack of active trading in the instruments. As our Accrued Interest author hints at, almost none of these instruments are liquid except by "accident," and how do you "mark-to-market" an asset that hasn't traded in months, or at all since you acquired it?
Worse, the liquidity that pours universally into these instruments (both the risky and less risky debt instruments) from the same sources for the same purpose has the effect of correlating them "artificially." The effect on portfolio diversification (even within a given CDO reference portfolio) is predictable. This looks very much like the recent effect on gold. If enough people seek diversification and efficient frontier advantages via investment in the same risky/non-risky asset baskets, the diversification (and efficient frontier) effect is lost as the assets become more and more correlated. This effect, unfortunately, only reveals itself when the correlation is most needed. Clever that.
What are the lessons for the Going Private reader? The combination of unaligned incentives (both on the intra-institutional and inter-institutional level), overheated first order liquidity (cheap cash), fickle second order liquidity (as distinguished from true liquidity), and a time lag between liquidity supply and performance feedback for that liquidity, should present strong, structural arguments that, when carefully examined, result in legions of smug looking Going Private readers sunning themselves while floating on liquid(ity) in their new yachts.