In the past two years, hundreds of lawsuits have been filed in the subprime mortgage/credit crisis. Like any broad category, the subprime-related suits break into subcategories, but one common feature they have is their complexity. Often, the pleadings alone are hundreds of pages.

This complexity means a large amount of electronic information, so sound search and cull strategies have never been more important.

You can cull at collection, after collection, during processing or after loading onto a review platform. Suit yourself. “When” doesn’t matter as much as “how,” because the volume of electronic data from the structured finance division of an investment bank is the size of a galaxy. You need it the size of a constellation.

Several federal rulings in 2008 state that expert advice should back up keyword searches. Counsel and litigants should develop those search strategies with consultants who know something both about the substantive subject matter, and about the methodology of keyword search testing and refinement.

There isn’t one standard off-the-shelf list of searches for all subprime litigation. The lexicon will depend on what subcategory you’re in.

The subcategories are not a simple list. Picture a three-dimensional cube with three axes: height, width and depth. The intersection of those axes determines the subcategory and the keywords to use. One axis is the parties. The second axis is the type of proceeding. And the third is the substantive nature of the claims and defenses.

The parties axis requires understanding the players: mortgage originators, real estate appraisers, retail banks, investment banks, commercial banks, bond insurers, bond rating agencies and investors, both institutional and individual.

The type of proceeding relates to the form and statutory basis for the claim. Is it litigation or is it a regulatory investigation? We all know how different the rules are between these two. In the federal- and state-level investigations now under way, it’s a given that targeted organizations divulge whatever they’re asked for. Then the investigatory agency is challenged with the deluge.

In one recent case, an agency received more than a million electronic documents. The first stab at keywords reduced it to a half million. Several more iterations with some consulting brought it down to 200,000, including 50,000 that would have been previously missed.

Many litigation cases are class actions, alleging violations of the Securities Act of 1933 and the Securities Exchange Act of 1934. Others are derivative lawsuits, where shareholders nominally take the reins of the corporation and sue for a right that the corporation could itself sue for, but won’t.

If it is a class action, it needs class certification. If it is an action alleging fraud and misrepresentation under the above acts, it will have to survive a motion to dismiss, commonplace since the Private Securities Litigation Reform Act of 1995, the “PSLRA”. In both instances, traditional discovery is stayed until these are determined.

And of course, the third dimension is the case’s substantive subject matter.

The complaints of the class actions and derivative lawsuits often focus on share value declines, or conversely, artificially high share values at the time of the plaintiffs’ purchases, which the plaintiffs allege are the fault of the company and/or its directors and officers because of mismanagement and misrepresentation.

Another issue is the “take it back, it’s gone bad” action, where an issuer of a Residential Mortgage-Backed Security is sued to repay what it received for a packaged bundle of subprime mortgages, because of deterioration within a set time limit.

“Gatekeeper” claims are suits against auditors or bond-rating agencies.

For example, the investment adviser’s claim: “You advised us to buy these bonds because ‘They’re backed by residential mortgages. What could go wrong?’” Another unique subcategory involves investors suing their advisors for giving their money to Bernard Madoff.

This list is not exhaustive. Other types of proceedings include ERISA lawsuits or actions for risk misrepresentation by underwriters issuing these collateralized debt obligations.

If an institutional investor sues an investment bank (party axis) in a securities class action (type of proceeding axis) for misrepresentation in offering statements in a CDO bond issuance (substance axis), the keywords list would include words like “super-senior,” “tranche,” and “grade.”

If a mortgage originator is sued in a class action for predatory or discriminatory lending, the list will include words like “FICO,” “teaser rate,” and “no-docs.”

If a non-class investor sues a bond-rating agency for allegedly dressing up garbage as gold, the list might include words like “risk” and “model.”

Look what that search might turn up. In 2007, two bond-rating agency analysts engaged in this instant message exchange:

Analyst 1: By the way, that deal is ridiculous
Analyst 2: I know, right. The model definitely doesn’t capture half the risk.
Analyst 1: We should not be rating it
Analyst 2: We rate every deal. It could be structured by cows and we would rate it.
Analyst 1: There is a lot of risk associated with it. I personally don’t feel comfy signing off as a committee member.

This was read aloud at a congressional hearing in Oct. 2008. On Feb. 18, 2009, it was quoted by a federal judge denying a defendant bond-rating agency’s motion to dismiss In re Moody’s Corporation Securities Litigation.

The nugget was found in the investigation preceding the early-stage motion to dismiss. A lot of rocks had been turned over by then, long before the “official” discovery stage of litigation.

The PSLRA’s pleading standards are beyond our scope here, but they guarantee the first battle in cases where it applies will be a motion to dismiss, with an automatic stay of discovery.

The stay, however, doesn’t mean you should hit the snooze button: the complaint, the motion to dismiss and the opposition thereto all require a detailed knowledge of the facts and documents proving them, from internally available sources and external research. This is early case assessment, or “pre-discovery discovery.”

To me, more actions now survive the motion to dismiss than they did a few months ago, and this might be the start of a trend. Two decisions in the Central District of California in late 2008–Countrywide and New Century–are notable. Although the rulings’ significance lies beyond the scope of this article, it is important to note that, as more motions to dismiss are denied, more of these massive cases will survive to the official discovery stage. Then, the floodgates will truly open.

The cost of resolving all this should not be driven up by unduly broad, unfocused discovery. It has already cost us plenty.

Active in litigation support and e-discovery since the late 1980s, Cliff Shnier is an attorney and independent electronic discovery consultant based in Scottsdale, Arizona. He has also owned a service bureau and held senior executive positions with national e-discovery providers. A graduate of the University of Toronto Faculty of Law, Cliff actively practiced law for 11 years and has extensive courtroom experience litigating complex commercial matters, as well as negligence and criminal cases. E-mail him at