Final Thoughts on Tavakoli's Structured Finance Classic
Book: Structured Finance and Collateralized Debt Obligations | Author: Janet M. Tavakoli | Publisher: John Wiley & Sons (2008) | ISBN: 978-0-470-44344-6
Twenty-six posts later, we’re done with this one.
I started this series partly out of curiosity and partly out of a conviction that the 2008 crisis has been explained badly in most places. Too much narrative, not enough mechanics. Too many villains and heroes, not enough honest accounting of how structures actually worked and who knew what.
Tavakoli’s book gave me what I was looking for.
What This Book Is
It is a technical manual that also functions as an investigation.
Most financial textbooks explain how instruments work in theory. Tavakoli explains how they work in theory, then explains how they actually get used, then explains the specific ways they get abused, then explains why smart people either missed this or chose not to see it.
That last part is rare. Most authors stop after the technical explanation. Tavakoli keeps going. She describes specific conversations with regulators who talked themselves out of caring. Specific deals with subordination structures she considered insufficient. Specific products she declined. Real-time accounts of ratings shopping, of mark-to-market manipulation, of fee income misrepresented as credit risk returns.
The book’s preface is worth reading on its own. She writes that “the irony of the complex CDO market is that the basic principles of sound finance are often violated in ways the models cannot capture. Models are a secondary overlay in determining fundamental value.” She then adds: “If you don’t know where you are going, writing a model isn’t going to get you there.”
That framing runs through the entire book.
The Chapters That Stood Out
Chapter 5 on CDO tranching lays the foundation. If you have never thought carefully about how subordination works, how cash waterfalls flow, and how rating agency attachment points are actually determined, this chapter will change how you read every financial product prospectus you ever look at.
Chapter 8 on mortgage loan abuses is the most harrowing chapter. Tavakoli documents the specific loan products that were being originated at scale: NINA loans (no income verification, no asset verification), SISA loans (stated income, stated assets), 40-year ARMs with negative amortization, option ARMs where the principal balance could increase. These were not edge cases. They were mainstream products originated by the largest mortgage lenders in the country and sold into securities with AAA ratings.
Chapter 13 on super seniors is the most technically surprising. The fact that 80 to 90 percent of the CDO structure had no standard definition, no standard pricing method, and no independent mark-to-market is genuinely shocking. The book was written years before this became a crisis, and Tavakoli had been flagging it in writing since 2002. She asked Moody’s about it. She asked S&P. She asked regulators. Nobody acted.
Chapter 14 on correlation should be required reading for anyone who uses quantitative models to make financial decisions. The core argument is simple: correlation is the least important of the three variables that drive credit losses, and yet the industry built massive infrastructure around modeling it. The models were unstable. The hedges based on them didn’t work. And the opacity of the model parameters made it easy to manufacture synthetic income. Tavakoli says spending zero time on correlation and more time on default probability and recovery rate estimates would have been a strict improvement. Based on what happened, she’s right.
Chapter 17 on the credit crunch brings everything together. She had made specific public predictions in 2007 about subprime losses and monoline capital shortfalls that turned out to be far more accurate than the projections from the Fed, from Citigroup, from CSFB. Her estimates of needed capital for FGIC, MBIA, and Ambac were higher than the rating agencies’ but aligned with what ultimately transpired. She was not guessing. She was doing careful analysis with better assumptions.
How It Holds Up
The book was written during the crisis. The second edition was completed in early 2008. Tavakoli was watching events unfold in real time and updating chapters as she went. Some sections have explicit dates in the text – “as of January 2008” – which gives the book a documentary quality that adds to its credibility.
The specific products she discusses – BISTRO transactions, CDO-squared structures, CPDOs, SIV-lites, leveraged super seniors – are partly historical now. You will not encounter these exact structures at their 2006 peak volume.
But the underlying patterns are not historical at all.
Rating agencies still have conflicted incentives. Regulators still lag behind financial innovation. Senior bank managers still often do not understand what their trading desks are actually doing. Complex products still get sold to investors who cannot independently price them. The revolving door between regulators and financial institutions still creates structural conflicts.
The tools Tavakoli uses to identify problems – follow the cash flows, understand who bears what risk and at what price, verify whether documentation matches the pitch, check the mark-to-market methodology, demand Moody’s tranching rather than the more lenient alternative – remain completely valid. They will remain valid regardless of which specific instrument is in fashion.
Who Should Read This
If you work in credit markets, risk management, structured products, or asset management, read it. The foundational material on CDO mechanics (Chapters 2 through 7) is some of the best accessible coverage of how these structures actually work.
If you work in investment banking compliance, regulatory affairs, or financial law, read it. The chapters on documentation, structural protections, and what can go wrong in legal interpretations are directly useful.
If you are a retail investor who wants to understand why your pension fund or bond fund holds things you’ve never heard of, read it. You probably won’t work through every technical section. But the preface, Chapter 8, Chapter 13, and Chapter 17 will give you a clear picture of what happened and why.
If you are just curious about how financial systems fail, read it. The combination of institutional incentives, model dependence, regulatory capture, and human nature that Tavakoli describes is not unique to CDOs. The same dynamics appear in any sufficiently complex market where measurement is difficult and the people doing the measuring have something to gain.
One Thing I Want To Emphasize
Tavakoli is not anti-finance. She ends the book by saying she is “an enthusiastic proponent of structured financial products.” She wants these markets to work well. The point of the book is not that CDOs are bad. The point is that specific structures, specific incentives, and specific failures of analysis and oversight turned potentially useful products into instruments of massive harm.
That distinction matters. The instinct after a crisis is often to ban the thing that caused the crisis. Tavakoli’s instinct is to understand what went wrong precisely enough to fix it. That requires more rigor, not less market activity.
She wanted investors to be sophisticated enough to protect themselves. This book is an attempt to make that possible.
This is the final post in a 26-part series on Structured Finance and Collateralized Debt Obligations by Janet M. Tavakoli.
Previous: The Future of Structured Finance: Regulatory Failure and What Comes Next