Jan 15

Subprime layers and the ABX Index

by David Harper, CFA, FRM, CIPM


FRM |

The CFA Institute today published a worthwhile 20-minute podcast on "The Pricing of Illiquid Securities in the Subprime Space" by Christopher C. Finger, Head of Research at RiskMetrics.

Two takeaways, for me. First, the three-layer framework isolates on a challenging aspect of structured security valuation. Second, he succinctly explains why the ABX helps (just helps) to value subprime securities.

More layers slows the information flow

On the three-layer framework. This is a common but convenient way to think about structuring: as the "distance" between asset and investor. If we are an investor in subprime, we can be "near the asset" with our direct ownership (e.g., the originator). Or at the other end, in a CDO, we might be two layers removed (or more, of course). Going from near to far, for example:

  1. Ownership of credit-sensitive assets: This includes mortgages, auto loans, and trades receivables. These are the underlying assets. In the case of an individual mortgage loan, cash flow is produced in the form of principal and interest (P&I).
  2. Asset-backed securities (ABS). A first-order pooling of assets (and precursor to CDOs). In the case of mortgage-based security (an MBA is a flavor of ABS, where the assets happen to be mortgages), pools of mortgages collateralize securities issued to investors. As a pool, the mortgages collectively create principal & interest (P&I). Investors receive this "waterfall" of cash flows. Investors can receive cash flows on a ratable, proportionate (single-class) basis or in a more tailor assignments (multi-class).
  3. Collateralized debt obligations (CDO). A second-order pooling of credit-sensitive assets. If the collateral issued to investors is itself an asset-backed security (MBS), this is really a re-structuring or a re-securitization (Culp sometimes calls this a reinsurance): the securities issued to investors are collateralized by asset-backed securities. The original assets are now "further away" in the structure. An additional layer of distance between investors and cash-generating assets.

Complexity as distance from the asset

Finger points out that keeping track of structured pieces, in addition to the underlying mortgages, is a "huge operational cost." He says it is not the math that kills you, but that when you are dealing with layers of structures, it is the speed of information. Structured subprime pricing is opaque because:

  • Banks don't have enough (market-based) price-discovery information
  • The securities are complex, but the complexity is due to many moving parts (more than the math)

Regarding price-discovery, ABX is the worst index. And the only.

The ABX is an imperfect but nonetheless useful proxy for the value of subprime mortgage securities. Note that a proxy does not need to be a pricing methodology, it can be "merely" a valuation reference or benchmark. And the ABX is not really an index, but rather a standardized portfolio that gives a window into contract prices for a sample of the subprime market. Each index series contains 20 MBS deals.  The index performs price discovery (what is the price of a contract that reimburses for default-related losses?) on only five subprime credit quality tranches (four series have been issued; each a block of time). Some fairly cite tranche paucity as a fatal weakness.

But the ABX index is useful because it's the one place where there is a liquid, price-revealing instrument for subprime-backed securities. As Finger notes, if these [the bonds in the ABX] are the most liquid bonds available, the ABX is a "pretty good indication of the market's sentiment around subprime MBS paper." So, again, he's not advocating the use of the ABX as a valuation methodology unto itself, but as means to incorporate the market view.

He notes that pricing can be approached either bottom-up or top-down (not mutually exclusive). Bottom-up is a granular analysis that weighs both the bonds and the structuring mechanism. Somebody must conduct a bottom-up analysis. But even for a plain-old MBS, this is tough work. Top-down refers to incorporating the market's general appetite for this type of investment (subprime).

The implication is that, while somebody must do the former (bottom up), you can't entirely omit the latter (i.e., the market's latest appetite for the asset). That is sound finance: no asset has value absent a market context. Intrinsic value is infected by subjective, temporal market views. An asset's value is a function of risk premiums, those risk premiums (e.g., equity risk premium) are not really, perfectly knowable in the moment, but they surely exist. Might as well take a stab.

Using the ABX like real estate comps

He likens the use of the ABX to the use of comparables (residential "comps") in the appraisal of residential real estate. That is, you check the comps, then you make adjustment for the differences (e.g., our house has a pool and additional square feet; our portfolio has longer duration).

So, of course you have to analyze the fundamentals of the actual holdings, but if you are trying to get an overall sense of the market's overall appetite and the volatility around these instruments, the index is fine proxy.


Comments

  1. Be the first to leave a comment!

Leave a Comment