Risky Business: The Real Economics Of Contingent-Fee Litigation In The Delaware Court of Chancery

Adam Pritchard is a law professor at the University of Michigan who is acting as a paid expert for one of the parties objecting to the fee award in Tornetta v. Musk. See Affidavit of Adam Pritchard, Tornetta v. Musk, 2018-0408-KSJM (Del. Ch. May 31, 2024) (Trans. ID 73297831) ¶1. (Disclosure: My firm had the privilege of providing pro bono representation to Professor Charles Elson who submitted amicus briefs in support of Tornetta in that matter).

As the Delaware Supreme Court considers that appeal, Pritchard  has released an article, Is Delaware Different? Stockholder Lawyering In the Court of Chancery, authored with two academic colleagues (Jessica Erickson and Stephen Choi). Their article concludes that “plaintiffs’ attorneys in Delaware are overcompensated—on average—for the risk of not getting paid for their work.”

Pritchard and his co-authors begin by assembling “data from every derivative suit and class action involving a public corporation filed in the Delaware Court of Chancery between 2017 and 2022.” From that sample, they calculate that “54% of the cases resulted in a fee-paying outcome for the plaintiffs’ attorney.” And from this, they “compute an overall implied multiplier of 1/0.54 or 1.87,” suggesting that “if plaintiffs’ attorneys can expect to receive fees in 54% of their Delaware cases, the attorneys need a multiplier of 1.87 to compensate for the risk of no fees.”

Pritchard and his colleagues then calculate that, in cases in their sample where the Court of Chancery awarded a fee, the mean “actual” multiplier was 2.59. Because 2.59 is larger than 1.87, they conclude that plaintiffs’ lawyers are “overcompensated … for the risk” they are taking and that “[t]his excess return comes out of stockholders’ pockets[.]”

That analysis, published in abbreviated form here, has been widely covered and promoted by those who seek to unshackle controlling stockholders (primarily, tech founders like Musk as well as private equity sponsors) from their obligations to minority investors by further reducing the extent and effectiveness of private enforcement of Delaware’s robust fiduciary-duty regime.

As one colleague in the stockholder bar noted, however, the article demonstrates a dramatic lack of “understanding [of] basic concepts underlying the contingency firms’ business models” and “many of their assumptions would quickly deteriorate if they spent five minutes speaking to someone who actually does this for a living.”

Every single step of Stockholder Lawyering’s analysis is mistaken.

A. Stockholder Lawyering Gets Basic Math Wrong

The math at the heart of the analysis is badly wrong and obviously so.

This all begins with a factual mistake. Pritchard and his co-authors state that “[w]e compute the fees per hour as equal to the attorney fees plus expenses divided by the number of hours. We utilize this definition of fees per hour because many attorney fee motions in Delaware do not separate attorney fees and expenses.”

Pritchard and his co-authors seem to be unaware that attorney fee applications in Delaware always include a breakdown of fees and expenses in the supporting affidavits, even if the expense figure is not provided in the motion itself. In a far more thoughtful, and nuanced recent paper, Prestipino and Klausner analyzed a larger set of Delaware fee awards—for “all derivative and class action lawsuits filed in the Delaware Court of Chancery between January 1, 2013, and December 31, 2023, involving publicly held corporations.” Prestipino and Klausner did it properly and excluded expenses from their calculation of fees.

Pritchard and his co-authors made a critical error in failing to do so.

Consider a simple example. Let’s assume that Law Firm is deciding whether to litigate a case with potential damages of up to $1 billion on contingency. Law Firm believes that prosecuting the action to the point of a recovery will require its lawyers to devote approximately $10 million in lodestar and advance $2 million in expenses. They believe there is a 54% chance of success.

Because there is a 54% chance of success, Prichard and his colleagues would say that an “implied multiplier” of 1.87 would fairly compensate counsel for contingent risk. They include the reimbursement of expenses as a “fee” in their calculation of the multiplier, so they would treat an “all-in” award of $18.7 million ($10 million * 1.87) as appropriate and any award above that amount as providing an “excess” multiplier.

But in the hourly scenario, Law Firm is paid $10 million win or lose and is not required to advance any expenses. The ex ante expected value is thus $10 million:

In the contingent scenario, however, Law Firm has a 46% chance of ending up in the Loss scenario, thus seeing its work go unpaid and losing the $2 million in expenses that it has advanced. In the Win scenario, the first $2 million of the $18.7 million all-in award simply pays off the expenses, so the true fee is only $16.7 million. Thus, the ex ante expected value of taking the case on contingency for a 1.87 multiplier is only $8.098 million—substantially less than the hourly approach:

To make the ex ante expected value of contingent litigation equal to the $10 million expected value of hourly representation, the all-in award would have to be at least $22.22 million (i.e., $20.22 million after deducting the $2 million in expenses):

Pritchard and his co-authors, however, would calculate a $22.22 million “all-in” award as a 2.22 multiplier (i.e., $22.22 million award divided by $10 million in lodestar = 2.22X). They would say that the difference between that multiplier and their 1.87 “implied” multiplier was “excess.”

This basic arithmetic proves obvious errors. And because Pritchard and his co-authors do not make their underlying data public, it is impossible to estimate how significant the effect of the error is. Their entire analysis would have to be redone.

B. Stockholder Lawyering Underestimates Risk By Excluding Common Outcomes Where Contingent Counsel Go Unpaid

Pritchard and his co-authors make other errors that cause them to dramatically understate the risks that plaintiffs’ counsel face. They “exclude[] from [their] review … books and records cases under DGCL § 220, and … filings that were subsequently consolidated with other cases.” From that sample, they determine that “54% of the cases resulted in a fee-paying outcome” and conclude from this that there is a “54% … probability of a fee-paying outcome.”

These decisions mean that Stockholder Lawyering is excluding four significant categories of outcomes that regularly result in plaintiffs’ law firms incurring hundreds of thousands or millions of dollars in lodestar for no compensation.

1. Fruitless 220 Investigations. In Delaware practice, a representative action is invariably preceded by months of pre-suit investigation using the DGCL’s books-and-records provision, Section 220. Section 220 investigations—which can involve the production of hundreds or thousands of documents—can be enormously time-intensive and require millions of dollars in lodestar. Yet many Section 220 investigations never result in a plenary action being filed. Plaintiffs’ lawyers will often review the documents produced in response to a Section 220 demand, determine that they show there is no viable case to be pled and move on without filing a plenary action. Even where the documents do show wrongdoing, other developments may prevent a plenary complaint from being filed, particularly where the underlying plenary claim would be derivative.

2.  Lost Leadership Disputes. Pritchard and his co-authors make a similar error in conflating a finding that “54% of the cases resulted in a fee-paying outcome” with an ex ante “54% … probability of a fee-paying outcome” for a law firm considering whether to pursue a contingent representation.

This is a significant error. The fact that one plaintiffs’ firm has a “fee-paying outcome” does not mean that all of the plaintiffs’ firms that pursued the case had a “fee-paying outcome.” To the contrary, in almost all of the most significant cases, there will be a leadership fight. The law firms that win the fight may get a fee-paying outcome at the end. But the firms that lose will go home empty-handed.

In Delaware, the Court gives significant weight to factors—the quality of the pleadings and counsel’s performance in the litigation to date—that incentivize plaintiffs’ firms to make a substantial up-front investment in developing their initial complaints before knowing whether they will be appointed to lead the action. As a result, the Court is usually asked to resolve leadership disputes where two or more competing groups have each drafted lengthy complaints of 100 pages or more that reflect the fruits of extensive, and time-intensive, pre-suit investigations using Section 220. The successful applicants may ultimately obtain a “fee-paying outcome.” But the losing applicants will not.

3. Partial Settlements. As explained above, Pritchard and his co-authors treat any case that includes a “fee-paying outcome” as a win for plaintiffs’ counsel and calculate the “actual multipliers” as of the time of the fee application. While they do not expressly explain how they account for partial settlements, it appears that they simply ignore whatever happens after the initial settlement. This means excluding a frequent category of outcome—a partial settlement with some defendants followed by a loss against others—in which plaintiffs’ counsel perform significant unpaid work that Pritchard and his co-authors completely ignore.

Consider three recent examples:

• In In re Tesla Motors, Inc. Shareholder Litigation (“SolarCity”), 12711-VCS,  the plaintiffs reached a partial settlement with most of the defendants (Tesla’s outside directors) in July 2020. The remaining defendant, Elon Musk, did not settle. Plaintiffs’ counsel then spent another three years litigating against Musk, losing at trial and seeing that verdict affirmed on appeal in 2023.
• In In re Columbia Pipeline Merger Litigation, 2018-0484-JTL, the plaintiffs reached a partial settlement with two of the defendants (the Company’s CEO and CFO) in June 2022. The third defendant (the buyer, TransCanada) did not settle. Plaintiffs won at trial with the Court of Chancery concluding that TransCanada had aided-and-abetted breaches of fiduciary duty by the settling defendants. In June 2025, the Supreme Court reversed and directed the Court of Chancery to enter judgment for TransCanada.
• In In re Straight Path Communications Consolidated Stockholder Litigiation, 2017-0486-SG, plaintiffs reached a small settlement with one defendant in August 2022. The remaining defendants did not settle and the case proceeded to trial. After trial, the Court concluded that the lead defendant had engaged in a “flagrant breach of duty” but that plaintiffs were entitled to only nominal damages. The Court subsequently rejected plaintiffs’ motion to shift fees to the defendant found to have flagrantly breached his fiduciary duties and held that plaintiffs’ counsel were not entitled to any fee. As of November 16, 2025, plaintiffs’ appeal is pending before the Delaware Supreme Court.

In each of these instances, it appears that Pritchard and his co-authors cut off their analysis at the time of the initial settlement. That means that in each case, their analysis ignores three years’ worth of uncompensated work by plaintiffs’ counsel after the initial settlement.

4. Appeals. Finally, Pritchard and his co-authors do not appear to adequately account for the fact that fee awards are sometimes appealed (either as an appeal of the fee award itself, an appeal of the merits ruling providing the basis for the fee award, or both). This results in them overstating the multipliers awarded because plaintiffs’ counsel sometimes have to spend additional time defending their fee on appeal and Pritchard and his co-authors do not factor in any of that time when calculating the “actual” multipliers awarded.

Each of these errors cause Pritchard and his co-authors to underestimate the risks of contingent litigation. In combination, these errors lead to gross underestimation.

C. Stockholder Lawyering Gets The Incentives Wrong

Finally, on a more fundamental level, Pritchard and his co-authors are asking the wrong question. They frame the inquiry as an effort to determine what level of “fee awards are necessary to incentivize lawyers to represent stockholders.” They suggest that any award in excess of that figure is “simply … giving away stockholders’ money without the stockholders receiving any commensurate benefit.”

This is flatly incorrect. As Prestipino and Klausner correctly observe, stockholders would want a fee-award regime that will maximize their expected recovery. This requires an analysis of the incentives created throughout the case to avoid incentivizing lawyers to settle cases for less than their expected value. But an approach that sets fees at the minimum level necessary to incentivize filing the case will also incentivize lawyers to settle cases for less than their full value, thereby harming stockholders and benefiting the controllers who have exploited them.

Let’s return to the hypothetical above. Law Firm is working on contingency and has incurred $10 million in lodestar and $2 million in expenses. The case has been tried and post-trial briefing has been completed. The Court has just heard post-trial oral argument and promised to issue its post-trial ruling in 90 days.  Law Firm believes that it has a 54% chance of prevailing at trial (and through appeal). If Law Firm wins, the damages will be $1 billion. Law Firm believes there is a 46% chance of an outright loss. The expected value of the case is thus $540 million.

As the parties leave the post-trial oral argument, defense counsel pulls aside Law Firm’s lead partner and offers to settle the case for $200 million.

What happens next?

The stockholder class would, obviously, want Law Firm to reject that lowball offer. And under a regime where Law Firm believes it will be paid a percentage of the recovery (say, 25%, plus expenses) with no lodestar cap, Law Firm is incentivized to do just that. Under that regime, settling has an expected value to Law Firm of $50 million. The expected value for Law Firm of pressing forward to the end is $134.1 million. The decision is, appropriately, a no-brainer.

But what happens in the world that Pritchard and his co-authors advocate for? They write that “the Delaware legislature might set a multiplier cap that exceeds 2.5, perhaps matching Texas’s multiplier cap of 4 or even going a bit higher to 5. We are hard pressed to see any justification, however, for double-digit multipliers.”

So, let’s imagine that the fee cap is set at 5X lodestar. Let’s imagine further that the Delaware General Assembly is more thoughtful than Pritchard and his co-authors and does not treat the return of expenses to be a “fee” subject to the cap.

In this world, the incentives all point in the wrong direction. Law Firm has $10 million in lodestar, so it has the same expected value from a $200 million settlement: $50 million. But now Law Firm’s expected value from pressing forward to the end (and thereby maximizing value for the Class) is substantially less. In the scenario where Law Firm succeeds, its fee will still be capped at $50 million because it has only $10 million in lodestar. But there’s only a 54% chance of getting any fee in that scenario, so that is an expected value of just $27 million. There is a 46% chance of a loss, where Law Firm loses the $2 million it paid in expenses (an expected value of -$0.9 million) and earns $0 in fees. So the total expected value to Law Firm of going forward is only $26.1 million versus an expected value of $50 million for settling. Law Firm will be heavily incentivized to settle and, if it does so, it will cost the class hundreds of millions of dollars of expected value.

This is a transparently awful incentive structure that, if adopted, would encourage lowball settlements undercompensating stockholders in the largest and most valuable cases.

For many of those advocating for a multiplier cap, this is a feature, not a bug. They should be honest about that goal rather than pretending that this is about protecting stockholder interests.