Friday, May 20, 2016

Notre Dame Deception Roundtable, part 4

Session 4 – Contracts and Securities
Discussion Leaders: Greg Klass, Ann Lipton, Andrea Matwyshyn
Matwyshyn: there’s a duty to perform in good faith in the US, but no duty to negotiate in good faith. If you have an integration provision, conversations leading up to the contract will be excluded from contract interpretation. If we let people lie leading up to the contract, what are we showing about our values and also about differences b/t our contract law and EU, where lies in negotiations may be actionable.
Klass: integration clause won’t prevent a defense of misrepresentation in negotiations, or the tort of misrepresentation.  Good faith is really interesting, but there’s a separate issue in contract law, the economic loss rule, which will prevent a claim for the tort of deception in the performance. So if you lie about your performance of the contract, only breach damages are available. But a precontractual lie isn’t covered by the economic loss rule.
Silbey: Tort cares more about diffuse harms, even if it’s hard to make out a claim, than contract.  Contract is about freedom to contract and freedom from contract, while tort is a different species of social values law cares about.
Klass: you want to look at what work the doctrine of pre-contract good faith does in EU.
Matwyshyn: people have sued over term sheets successfully in the EU.
Klass: before contracting, your duty is not to misrepresent; you don’t have to look out for the welfare of the other party. You can fight for a larger share of the pie as long as you play by the rules.  Related question: can you contract out of fraud liability and say that lying is permitted in your negotiations?  Delaware Ch. Ct. case: the court says you may not do that.  There is an obligation not to lie that’s fundamental; but you can include in your contract a representation of no reliance, effectively precluding any action for misrepresentation.  In M&A, seller is often worried about misrepresentation claims, so they want the buyer to sign an agreement saying it’s not relying on anything outside the agreement. So all you need is the magic words.
Lipton: NY has done this w/sophisticated parties.  Mortgage-backed securities/synthetic CDOs contracting w/German bank.  Court said that it was misleading of the P because the P represented that it wasn’t relying on the D’s representations in the contract. 
Klass: Maybe one side understands the magic words and the other doesn’t, and instead of saying “there’s no fraud liability” which would be very clear they allow confusion.
Lipton: Securities is different b/c of the multiple disclosure obligations.  Deception rules then implement the disclosure obligations; it’s about setting up an information market, and thus it’s not just intentional deception that matters but the quality of information. Capital formation/market structure as well as consumer protection.
Klass: Buell comes to it as a federal prosecutor with generic anti-fraud statutes; that’s where he starts, and maybe his approach focusing on deception is more fit to those.
Lipton: there is a rather extensive system of claims that don’t require any showing of intent—either strict liability or negligence. Puffery piece by David Hoffman: his framework doesn’t work for securities b/c there’s no intention requirement so his proposal to allow Ds to rebut by showing no intent to defraud is not helpful.
Goldman: why the different rules for securities?
Lipton: because this is about capital formation.  May have started as consumer protection, but evolved to want a deep and effective secondary market for trading. You therefore need a standardized info package. Lots of products can’t be investigated and are in some ways interchangeable; there are debates about whether/why the market wouldn’t generate the info w/o requirements. Billions of trades a minute.
Matwyshyn: it’s all about trust. Risk of investing in low credibility securities and risks of playing poker sometimes aren’t that different in terms of the numbers.  Maybe we are fetishizing this area of the economy in ways we don’t others.
McG: because for these other historical reasons and purposes you have such an elaborate set of disclosures, changes the nature of what deception means.  Information you’re owed as a backdrop to define what deception is: there is rough consensus about what has to be disclosed and how.
Lipton: there’s now a circuit split on what fraudulent omission is.  The big antifraud statute, 10(b): whether it’s deceptive to fail to disclose required info under 10(b). If you have a background expectation it will be disclosed, 2d circuit says that yes, it’s deceptive; 9th Circuit says no, there has to be something affirmatively said.  She doesn’t see the 9th Circuit’s logic. SEC can definitely bring a claim in either case; the case law is muddled by a view about how much we trust private plaintiffs to bring these cases when no one actually read the documents b/c it’s all a fraud on the market theory anyway.
McG: gets it back to who are the right parties to sue—it might not be the people who are deceived.  Systemic problems stemming from deceptive omissions.
Lipton: fraud on the market is very much an injury to the market, not to heterogenous consumers. We are supposed to use objective reasonable person standard, but in fraud on the market courts look through the lens of sophisticated people and in calls to widows they look through unsophisticated, even though that’s not the formal doctrine.
Matwyshyn: classes of trusted intermediaries have special roles and liabilities in this regime.  [And that interacts w/puffery and falsity, b/c things that might be nonfalse if said by others can be false if said by them.]  Frank Pasquale: new tech means we lose some checks on intermediation we used to have, as w/sophisticated algorithms that engage in billions of transactions/minute.  All it takes is one problem and no one is auditing the code.  Historical example: Brokerages lied about completing trades in-house b/c they couldn’t keep up w/the market: ended up w/regulatory intervention, lots of closed brokerages.
Eric Goldman: interested in the idea that securities market needs all this regulatory structure to be trustworthy enough to proceed.  What distinguishes this from other markets, like the eBays of the world where reputation is enough to build a trust market?  They’re both pushing stuff.  Type I/Type II errors: people sue b/c stock price went down; he thinks that’s bad. Should be concerned about both types of errors. 
Lipton: Congress made it really hard to bring a securities fraud case right now; pleading requirements, discovery bar; Type II errors are really unlikely.
Goldman: shows you that a regulatory structure needs constant tweaking to avoid the pendulum swinging too far. What about securities led to us building that oversight?  Case study of too much regulation.
Lipton: eBay as a company has the ability to stand behind sellers. NYSE used to have ability to stand behind companies. Regulation means it’s less necessary to be on NYSE b/c I know you have met requirements that the SEC stands behind. If you come from another country w/lower securities laws and you list here, that sends a signal to investors that you’re more credible and you have lower cost of capital.
McKenna: it’s also systemic risk. If eBay goes down, it doesn’t take down the entire economy.  That’s why you care about structural features—runaway effects of a crash. Also explains more extensive reporting requirements: thicker info requirements.  If eBay goes away, you just have to buy stuff in stores, but you don’t get a Great Depression (it’s just depressing).
Silbey: these disclosures aren’t actually transparent.
Lipton: but computers can and do interpret them, and sophisticated people can look at companies and compare them across an industry, which helps in trading.
Silbey: aren’t they routinely scrubbed and managed?
Matwyshyn: some things you can’t scrub. You have to talk about material litigation, for example. Bird’s eye view into how the company sees itself.
Lipton: I was a plaintiff’s lawyer and I’m skeptical but even I think there’s information there.  Commodities disclosures are different.  Pages of boilerplate disclosures of risks. Earthquakes could affect Twitter. You may think this is useless, but it turns out that people do econometric studies and those risk disclosures do affect stock prices. Computers look for tiny changes in language, and differences are caught that way.  You can find accounting fraud by crunching numbers and looking at language choices. When people commit fraud, they use different language.
Matwyshyn: companies in same industry were talking about tech in very different ways. 2004: Google didn’t disclose risks of security breaches in the same way Microsoft did.  You can track learning in the industry. 
Lipton: standardized set of disclosures allow you to detect patterns, not even as extreme as detecting fraud, through human and computer review.  When companies have bad news, they use bigger and vaguer words. 
McKenna: this is very far to the end of the structural harm line. Also there are lots of mediating sophisticated parties, so disclosures can be more useful here than in privacy. Also more standardized, instead of “say whatever you want and you’re going to be held to it.”
McG: there is one standardized disclosure in privacy, and it’s financial.  You don’t have to use the standard form, but there’s a safe harbor, so everyone does. Computer scientists at CMU did a computer analysis of them, which is routine in the securities space, and came up w/lots of interesting observations about regional differences, and found some companies breaking the law by their own disclosures, etc.
McKenna: if disclosure is the means to regulate, then should we require standardized disclosures?
Lipton: that’s good, but also need capacity to actually read them, whether human or computer.
Said: so context sensitive: “promotional consideration furnished by” is standardized but doesn’t solve problem (if problem there is). Extent to which digital tools are worsening deception problems b/c of ability to scrape, use hidden info, unsettle expectations; but also digital tools may be part of solution, whether using info commons or to detect fraud.
Perzanowski: nothing stops requiring a disclosure to be effective.
Lipton: that works unless there’s a lot of heterogeneity—experts in securities help.
Klass: misrepresentation in contracts includes nondisclosure, but it’s a vague standard: reasonable/not disclosing violate goods faith. The only way it works is that you get repeat situations: if there’s termites in your house you have to disclose; if you’re an oil company you don’t have to disclose you know there’s oil on the farmer’s land.  His own take: common law of fraud/contract is that we have clear norms about affirmative lies, and law piggybacks on those; that handles new situations. We don’t have strong intuitions on failure to disclose.
McKenna: tort is riddled with uncertainties about acts v. omissions writ large.
Matwyshyn: real estate contracts are a good example: regulatory interventions to explain what you have to disclose. It’s cooperative set of regimes working together.  Theme of this session: the focus on methods of detecting deception and fraud when it’s happening.  Sec reg might be better at that than some other contexts. 
Lipton: clearly there’s a bunch of fraud; sometimes computers are easier to fool than people, as when there are fake merger announcements that computers think are real and people could easily detect as frauds.  However, there’s a lot of money to be made in early detection, so it also happens.
Said: In speech arena, we have lots of worry about chilling through misreading/understanding whether speech is false.  We haven’t talked about listening or interpretation here.
Klass: pitch for Grice and implicature. A rich theory about how we interpret not just literal but implied meanings, including irony. Cost-benefit analysis may be built in. That’s how a lot of the law of deception piggybacks on extralegal interpretive norms.
McKenna: this all sounds like duty to me. Affirmative misrepresentation v. failure to disclose—this is the difference b/t someone who’s begun to act and thus has the duty to do it reasonably well, versus when I never start and have no duty to continue.  Regulation can also create duty.
McG: sometimes untidiness in law is based on different interests being served by different silos, and we should be willing to be comfortable w/that.
McKenna: but we should be clear about what we mean rather than assuming it has a fixed meaning.
Klass: it’s not common purpose or justification, but that there’s a common set of design questions that repeats across different fields.  My way of thinking: most of those are contained in the common law elements.  In this area of law: what’s the deal w/scienter? What’s the deal w/reliance? 
Silbey: basic things are missing from TM that could be borrowed.
McG: do you want a scary regulator like the SEC?
McKenna: think about why features work in some areas and not others.
Lipton: law keeps the corporation and the stock certificate relatively stable in what they are, so they’re relatively interchangeable. B/c of relatively homogeneous set of products, it’s easier to regulate them.
Matwyshyn: it takes a “river on fire” moment to have a meaningful evolution.  If we look historically at when quality control has meaningfully improved, what would it take to create change in deception regulation?  [FDA: it took a lot of dead kids.]
Gadja: news sites shifted from anonymous comments to Facebook in part b/c of all the defamation.
Lipton: scandals also produce incremental responses. Bork issue = just video rentals protected.  Harder to get overarching response, in the US.
McG: though other countries have done it.
Silbey: dilution added to TM act as a response to market forces.
McKenna: Even the SCt has no way of thinking about how to reconcile these different fields, as 1A discussion showed.  Alvarez is totally unsatisfying about why SVA is unconstitutional but TM is just fine.  Modern TM law is nothing like history or tradition was, which is why their explanation was wrong.
Silbey: Alvarez was not about TM.
McKenna: they think TM law is totally fine; they used TM to explain why the SVA was bad.
McG: Alvarez pro-US briefs tried to brief TM law as “uh-oh, be careful what you do so as not to destroy it.”  Thus the Court may have been trying to cabin the force of the opinion.
McKenna: gives us reason to think harder about the kinds of harms at issue. 

Silbey: in fundamental rights cases, the Court spends lots of time identifying the harms in fundamental rights like marriage cases. They seem unable to do so in these cases however.
McKenna: there’s so much assumption about what deception is.
Said: what if we looked for tolerated confusion/efficient confusion? 
Silbey: they had that in the briefing in Alvarez—a lot of discussion of the benefits of lies.   Flatness of discussion of variety of interests in IP cases, compared to the discussions of competing values in securities law etc.
Lipton: that’s a public choice issue—hasn’t been people with lots of money/big megaphone on the other side of IP cases.

No comments:

Post a Comment