Cash vs Synthetic Arbitrage CDOs: What's the Difference?
Book: Structured Finance and Collateralized Debt Obligations | Author: Janet M. Tavakoli | Publisher: John Wiley & Sons (2008) | ISBN: 978-0-470-44344-6
Chapter 9 does something really useful. It puts a managed cash arbitrage CDO and a managed synthetic arbitrage CDO side by side and walks through every step of building each one. Same deal size. Same goals. Very different mechanics.
This is where you start to see why synthetics took over the market.
Two Deals, Same Size, Very Different Structures
Both deals in the comparison are EUR 500 million. Both are cash flow deals, meaning the assets themselves pay back the liabilities. Both are investment grade.
But the tranching looks different from the start. The cash deal has a Baa3 average portfolio rating. The synthetic deal has Baa2. The cash deal has no super senior tranche. The synthetic deal has one, at 86.5% of the total deal. That difference is everything, and we will get to why in the next post.
The Arranger and the Manager
Before any portfolio gets assembled, a bank arranger needs a manager. What does the arranger look for? Track record, name recognition, portfolio managers who have been at the firm long enough to prove they are not just passing through. If the manager is willing to hold equity, that is a big plus. Prior CDO management experience is gold.
Managers are equally picky. They want to know their fee (40 to 50 bps per year for a cash deal, 12 to 25 bps for a synthetic), and they want to know how much of that fee sits senior in the waterfall. They also want strong distribution, especially for mezzanine and equity pieces, and they want the arranger to absorb warehouse funding costs.
Tavakoli lists some of the better-known names at the time: AXA, PIMCO, TCW, TIAA-CREF, Western Asset Management. Her note is sharp: “CDO managers from the ridiculous to the rare sublime have entered this lucrative area in droves.”
In the example deals, managers take 51% of the equity, which means the manager can exercise a deal call without needing other equity holders to agree. That is a structural detail with real consequences.
The Mandate Agreement
Once both sides are interested, they sign a mandate letter. It sets out deal economics, timeline, and usually gives the manager an exclusive for a fixed period. The arranger often works multiple deals at the same time. No problem, as long as they have enough people and trading capacity.
Assembling the Deal
Around the same time, portfolio selection starts. A law firm gets picked. Venue and SPE administration get sorted. The arranger starts premarketing, especially if equity and mezzanine need to find homes. Investor feedback can still change structural features at this point.
For the cash deal, document drafting includes a warehousing agreement. For the synthetic deal, there is no warehousing agreement. That alone tells you something about complexity.
CDS Language for the Synthetic Deal
For a synthetic CDO, the credit default swap language matters a lot. The goal is language that protects the credit protection seller, which is the SPE, which means the investors benefit.
Credit events are limited to three: bankruptcy, failure to pay (minimum EUR 1 million), and modified restructuring. There is also a EUR 10 million threshold for the default trigger. Deliverable obligations are senior unsecured debt, with up to 10% subordinated bank paper allowed. All payments are in euros, so investors take no currency risk.
Selecting the Portfolio
This is the most important part of any arbitrage CDO.
For the cash deal, the arranger’s trading desk has to actually find bonds in the market. Investment-grade deals can include up to 25% high-yield to make the arbitrage work. In Europe that is hard because high-yield issuance is thin, so managers often go to the US market. Most European investment-grade bond supply for decent spreads is concentrated in telecom and auto. Finding the right bond in the right size is genuinely difficult.
For the synthetic deal, credit default swaps can reference virtually any obligation. The arranger does not have to scour the market. Much easier.
There is a debate here that Tavakoli documents honestly. Some investors love managed deals because a good manager can avoid Enron or Railtrack before the rest of the market catches on. Others hate managed deals because they feel managers introduce “incompetence for a fee” or, worse, moral hazard when the manager holds equity.
The rating agencies are not much help in resolving this. Moody’s penalizes a deal if the manager has a bad track record, but will not give credit for a great one. That asymmetry frustrates a lot of structurers.
Rating Criteria and Restrictions
Rating agencies are not consistent with each other. An entity that S&P considers investment grade may not be investment grade under Moody’s. Fitch has its own classifications. And agencies are sometimes not even internally consistent across deals.
One practical consequence: you can sometimes arbitrage the rating agencies. For a synthetic corporate deal, S&P methodology might require less subordination to get to AAA than Moody’s. Pick the agency, save money.
For the cash deal: average portfolio WARF of 319 (between Baa1 and Baa2), minimum diversity score of 30, initial score of 33.8. Maximum industry concentration is 12%, with limits on country exposure and non-euro currency (max 40% USD). High-yield assets can go up to 25%, but default volatility for single-B assets is nearly ten times that of BBB assets. The manager has to barbell the portfolio.
For the synthetic deal: higher quality starting point with WARF of 236 (closer to A3). Minimum diversity score of 40. No corporate subordinated debt. No ABS. No emerging markets. Maximum country exposure 35%. CDS maturities start at five years from the deal close.
Tavakoli drops a jarring fact here. In 2006 Moody’s admitted that loss rates from BB non-investment-grade assets were the same as for BBB investment-grade rated assets. Their ratings had failed to distinguish between the two. Since some funds must sell non-investment-grade assets on downgrade, that distinction matters enormously. The data just did not support the boundary.
Substitution and Reinvestment
For the cash deal, there is a minimum spread requirement for reinvestment. Assets that mature during the reinvestment period get replaced with longer-dated assets. Recovered value from defaults also gets reinvested.
For the synthetic deal, the manager can substitute up to 10% of the portfolio per year with no limit for credit-impaired substitutions. Substitutions must be investment grade. The minimum spread test equals the average portfolio spread at inception.
Warehousing
No warehousing agreement is needed for the synthetic deal. The credit derivatives trading desk acts as the warehouse, and ramp-up is near-instant. Cash deals need two to six weeks of warehousing for investment-grade arbitrage CBOs. ABS-backed CDOs can need two to three months just to warehouse, because the collateral is hard to find and the analysis is deal-specific.
Warehouse funding and hedge costs in cash deals fall on either the manager or the arranger. That is a real cost the synthetic deal simply does not have.
Pricing and Closing
Pricing happens within a week of closing. At the closing date, the SPV takes legal ownership of the bonds. The SPV buys collateral at the arranger’s acquisition price.
Ramping Up
After closing, ramp-up is the period for acquiring remaining portfolio assets.
For cash deals, this takes four to six weeks for investment-grade deals, and potentially several months for ABS-backed CDOs. Some early highly leveraged loan deals failed to ramp up fully in the allowed time. The deals had to be scaled back. If portfolio diversity score or other parameters are not met, there can be penalties and partial note liquidation.
For the synthetic deal, ramp-up is complete at closing or within days of it. CDS traders can retain risk positions with confidence because the market is more liquid than cash bond markets, where a specific bond of limited issue size might just not be available.
Reinvestment Period and Noncall Period
Cash CDOs typically have a five-year reinvestment period. During that time, principal from maturing assets, defaults, and prepayments goes back into new collateral. The average life of the deal can shift during this period, so the documents specify acceptable ranges. The reinvestment period extends the average life and tends to increase equity returns.
The synthetic deal has no reinvestment period. It has a bullet five-year maturity. The manager trades to avoid losses or exploit gains, but there is no structural mechanism to reinvest principal.
Cash CDOs also have noncall periods, typically five years. Equity investors usually control whether the deal gets called. In the example, the manager holds 51% of equity and can call the deal unilaterally.
After the noncall period, tranches can be called at a premium equal to half the annual coupon. There is also a clean-up call provision: if a tranche balance falls below 10% of original due to amortization, it can be called at par because the servicing cost exceeds the arbitrage benefit.
Synthetic CDOs in this example have no call provision.
The next post picks up where this one leaves off: pay-down period, weighted average life, early termination, the super senior advantage, waterfalls, PIK tranches, and equity cash flows.
Previous: Bear Stearns, Hedge Funds, and the Mortgage Fallout Next: CDO Waterfalls and the Super Senior Advantage