Jesse Litvak, Singer of Tales.

Photographer: Douglas Healey/Bloomberg

When Can Bond Traders Lie to Their Customers?

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

Jesse Litvak is a former Jefferies bond trader who lied to his customers. Specifically, he traded residential mortgage-backed securities, and he would tell buyers of those securities inventive little stories about the prices he had paid for them. So for instance, a customer came to him and asked to buy some bonds, he described for the customer "a fictional back‐and‐forth between himself and an unnamed, non‐existent third‐party seller," and he ended by telling the customer that he'd paid $53.00 for the bonds. Then he charged the customer $53.25. The customer thought that Litvak had worked hard on his behalf and gotten paid a quarter for his efforts. But really Litvak had owned the bonds all along -- he'd bought them days earlier for $51.25 -- and made an undisclosed profit of $2.00.

The question is: Is this a crime? It might seem like lying to your customers would always be a crime, but that is not true at all. Lying to your customers is only a crime if it is fraud, and it is only fraud if the lies that you tell them are "material," and the lies are only material if there is "a substantial likelihood that a reasonable investor would find" them "important in making an investment decision." Lying to your customers about whether an investment is a Ponzi scheme is fraud. Lying to your customers about what color socks you are wearing probably isn't.

But lying to your investors about the price you paid for bonds is genuinely interesting. Is it fraud? It sounds like fraud. But when Litvak was arrested and charged with securities fraud in 2013, his defense was that his lies were not material, because no reasonable investor would care about them in making an investment decision. There are two main arguments for that:

  1. The only thing that a reasonable investor should care about is the value of the securities, not the price that Litvak paid for them. If someone thinks a bond is worth more than $53.25, he should pay $53.25 for it, regardless of whether Litvak paid $53.00 or $51.25 for it. The only information that should be material to investors is information about value (is the issuer bankrupt, etc.), not information about historical prices.
  2. Bond traders just lie all the time, so no one would ever believe anything a bond trader says, so obviously nothing a bond trader says can be material.

Those are both good arguments! They are also sort of ... gross? The federal judge in Connecticut who was in charge of Litvak's trial refused to let Litvak's expert witness (Ram Willner, a former bond manager with a Ph.D. ) testify about either of them. Without that testimony, Litvak was convicted of fraud and sentenced to two years in prison; he appealed, and today the U.S. Court of Appeals for the Second Circuit threw out his conviction and sent the case back for a new trial, this time with Litvak allowed to make those arguments to his heart's content.  I don't particularly fancy his chances of convincing a jury of either of them, based on the expert testimony of a finance Ph.D., but you never know. 

The first argument -- that information about what a dealer paid for bonds is not relevant to a reasonable investor who has done his own homework on value -- is sort of appealing, but probably not right. Here's how the court explains it (citation omitted):

Because RMBS lack an efficient, transparent secondary market through which value can be determined objectively, traders set the value of the security, and hence the price each is willing to accept as a seller or buyer, by engaging in “rigorous valuation procedures” involving the use of certain “analytical tools and methods.” Thus, firms trading RMBS rely upon sophisticated computer‐pricing models, often developed by professionals with applied‐mathematics backgrounds, to determine the subjective “value” of the securities. Certain testimony at trial supported a conclusion that this process, and a determination of the amount an investment manager is willing to pay for a security, nearly always takes place prior to the manager approaching a dealer such as Jefferies, here represented by Litvak, to negotiate the price of that security.

With such testimony before it, a jury could reasonably have found that misrepresentations by a dealer as to the price paid for certain RMBS would be immaterial to a counterparty that relies not on a “market” price or the price at which prior trades took place, but instead on its own sophisticated valuation methods and computer model.

On this view, smart investors in residential mortgage-backed securities -- and Litvak's customers were big smart investors -- don't just pay whatever their dealer asks for those securities. They have their own idea of what those securities are worth. There is ... math ... involved. They go to the dealer already knowing what the bonds are worth, and all that matters is whether the price the dealer charges them is less than what they think the bonds are worth. As opposed to stock investors, who have an efficient market, and who therefore just conclude that a stock is worth whatever the market says it is worth.  

That view of the stock market is obviously a bit idealized -- is a stock price really an objective determination of value? -- but the view of the RMBS market is no less idealized. I mean, sure, you math up the value of a bond, you decide that it's worth $55, and you go to a dealer hoping to buy it for less than $55. And if the dealer tells you it's trading at $53, you think, score, I can buy it for less than it's worth. But if the dealer tells you it's trading at $20, you don't necessarily think, score, I can buy it for way less than it's worth. You might think: Perhaps my math is wrong? Perhaps the market knows something I don't? The reasonable investor, even in mortgage-backed securities, doesn't decide what price she'll pay based purely on introspection. There is a market out there, with prices, and those prices might offer a clue about value. And so lies about those prices might deceive investors about value.

Or just more practically, whatever you believe a bond is worth, and however sure you are of your beliefs, you still might have to sell it. So knowing where the market is will always be of interest to any reasonable investor.

Still, you can see the appeal of this argument. If a bond investor is doing something socially worthwhile, it is by determining for herself what bonds are worth, and trading when prices diverge from her estimates of value. A big bond-fund manager who relies on the price a dealer gives her, without having her own independent idea of the bond's value, in some sense deserves to be cheated.

Litvak's other main argument -- that bond traders always lie, so no bond trader's lie could be material to a customer -- is less appealing, but more convincing. Here's how Litvak's lawyers put it, as quoted by the court (emphasis added ):

Where a manager follows rigorous valuation procedures, as was the case here, consideration of, or reliance on, statements by sell‐side salesmen or traders concerning the value of a RMBS or the price at which the broker‐dealer acquired it or could acquire it, are not relevant to that fund’s determination with respect to how much to pay for a bond. In Mr. Willner’s opinion, such statements from sell‐side sales representatives or traders are generally biased, often misleading, and unworthy of consideration in trading decisions. Accordingly, such statements from sell‐side sales representatives or traders are not material to a professional investment manager’s decision‐making.

Of course that is just Mr. Willner's opinion. But if it is a widely shared opinion among bond investors, then a salesman's lies about bond prices really can't be fraud, because no reasonable investor would believe them. This is the sense in which "everybody's doing it" really is a defense against fraud charges. To be fraud, lies have to be material, and if lies are so common that they are ignored, then they can't really be material.  It is a used-car-salesman defense, if you will.

This one is pretty controversial, so I will just leave it here. I'll note that the prosecutors offered testimony from several bond investors who said that Litvak's statements were material to them, so Willner's opinion may not be shared universally. On the other hand, some of that testimony itself supported the notion that bond trading is a war of all against all, fought mainly by lying, so who knows. And other cases similar to Litvak's make it seem like lying about bond prices is not an isolated phenomenon. It remains difficult to know for sure exactly how much dishonesty is expected, and thus allowed, in the market for residential mortgage-backed securities. It's a question for experts to argue over, and for a jury to decide.

Which is really weird! You could imagine a rule. The rule might say "it is always illegal for a bond trader to lie to his customers" (even about socks?), or it might say "it is always illegal for a bond trader to lie to his customers about the price he paid for bonds," or whatever. Congress, or the Securities and Exchange Commission, would write it down, and then everyone could know about it. Or you could have a different rule, like, "lie to your customers as much as you want about bond prices, it is their problem to figure it out." Then again everyone would know about it, and traders could lie. But customers would be on their guard, and would know to check prices with multiple dealers and punish lying dealers by not trading with them.

But now an appeals court has devoted 84 pages to this question and decided that there is no rule. Maybe lying to customers about bond prices is a crime, maybe it's not a crime, and it's up to 12 randomly selected laypeople to decide. Leaving the rules of financial markets to be determined based on after-the-fact review by non-experts chosen at random and empowered to impose prison time is a very American system. But boy is it strange.

I get the sense that the Second Circuit knows how strange it is. That's why it sent this case back for a new trial. In theory, a jury can decide whether or not Litvak's lies were a crime, but its decision made the Second Circuit uncomfortable, and it sent the case back for a different jury to consider, this time after hearing all of Litvak's arguments, even the distasteful ones. 

This seems like part of a broader discomfort, which you can also see in the Second Circuit's Newman insider trading decision last year. A lot of stuff that is widely accepted in financial markets rubs outsiders the wrong way. Sometimes hedge fund analysts talk to corporate insiders, for instance, and sometimes bond traders are less than fully forthcoming with their customers. Those practices develop over time, and are embraced by industry norms to a greater or lesser extent. And then some outsider, generally a prosecutor, discovers those practices and finds them distasteful. Being a prosecutor, he expresses his distaste by trying to imprison people. And since his distaste tends to be shared by a lot of laypeople, it is not that hard to get a jury to agree that these commonly accepted practices are in fact fraudulent and criminal.

As a system for expressing distaste, this works pretty great. As a system of financial regulation, or of criminal justice, it is perhaps not ideal. Congress or the SEC could reform the financial industry by explicitly deciding which norms need changing, and then writing new rules that tell people which lies, or which sorts of inside information, are forbidden. But they mostly don't, and leave it up to enforcement actions and juries to decide. And that keeps making courts nervous.

  1. The example, and the quotation, are from today's Second Circuit opinion in Litvak's case. The Second Circuit actually distinguishes three kinds of lies, which I have conflated a bit in the text:

    • "First, he misrepresented to purchasing counterparties Jefferies’s acquisition costs of certain RMBS" that he was selling out of inventory, for instance telling an AllianceBernstein fund that Jefferies had paid $58.00 for some bonds that it had actually bought for $57.50.
    • "Second, Litvak misrepresented to selling counterparties the price at which Jefferies had negotiated to resell certain RMBS" in riskless-principal trades. For instance, he "falsely stated to a representative of York Capital Management ('York'), a hedge fund that owned certain RMBS, that Litvak had arranged for Jefferies to resell those securities to a third party at a price of $61.25," and convinced York to sell them to him at $61.00. But his actual third-party bid was at $62.375.
    • "Third, Litvak misrepresented to purchasing counterparties that Jefferies was functioning as an intermediary between the purchasing counterparty and an unnamed third‐party seller, where in fact Jefferies owned the RMBS and no third‐party seller existed." That's where the example in the text comes from: Magnetar bid $50.50 for some bonds that Jefferies had in inventory (and for which it had actually paid $51.25); Litvak talked Magnetar up to $53.25 based on the fictitious conversation with the fictitious seller.

    These are all slightly different things but for our purposes they come to about the same place.

    Incidentally, I am saying that he "did" these things, not that he "allegedly did" them, because he was tried and convicted for doing them, and because he "didn’t dispute that he made misstatements." The dispute is not about whether he lied, but about whether those lies were material.

  2. That's $2.00 per $100 principal amount, or $14,000 more total than the customer thought Jefferies was making.

  3. Obviously nothing here is ever legal advice, you know the drill.

  4. I fell into a bit of a rabbit-hole wondering about hypothetical lies in this vein. So much of the business of finance is a people business, and who can know what an actual human investor will be looking for in his actual human salesman? Is lying about your golf handicap material to a reasonable investor? No, obviously, but might it be material to making an actual sale? Maybe?

  5. This is sort of a weak-form-efficient-markets argument.

  6. There's another expert, a lawyer named Marc Menchel, most of whose testimony was excluded, but Willner is of more interest to us. From the opinion:

    The first expert witness Litvak proffered was Ram Willner. According to Litvak’s expert disclosure to the government, Willner holds advanced degrees in business administration with a focus on finance, served as a professor at leading business schools, and gained “extensive experience in portfolio management in the fixed income asset class, including extensive experience in the analysis and purchase of Residential Mortgage Backed Securities” during his employment by, at various times, Bank of America, Morgan Stanley, PIMCO, and a hedge fund. 

  7. The Second Circuit also threw out some charges against Litvak, for fraud and false statements against the U.S. Treasury. These charges came from the fact that Litvak sold some bonds to counterparties who were managing "Public-Private Investment Funds" that included Treasury money, so the government argued that he was defrauding the Treasury. The Second Circuit disagreed, in 17 pages of analysis that I found persuasive but also astonishingly boring; let us never speak of it again.

  8. Then why do they trade? I love the phrase "an efficient, transparent secondary market through which value can be determined objectively." Courts are adorable. 

  9. The court underlined part of this quote for legal reasons that do not concern us; I have bolded a different portion.

  10. Litvak also made a related everyone-was-doing-it point, which the trial court also excluded and which the appeals court also reinstated. It's basically that Jefferies never stopped him from doing this stuff, so it must have been okay:

    Litvak sought to introduce evidence that, during the relevant time period, supervisors at Jefferies—including his supervisors—regularly approved of conduct identical to that with which Litvak was charged. The District Court characterized the proffered evidence as improperly “suggest[ing] that everybody did it and therefore it isn’t illegal.” But Litvak’s counsel did not proffer the evidence for that purpose and such an argument in summation could have been properly proscribed by the District Court. As Litvak’s counsel stated at trial, this evidence would provide “a fair basis upon which to infer that when Mr. Litvak did the very same thing, . . . the supervisors saw and approved of [it] as standard operating procedure.” Such an inference would support Litvak’s attempt to introduce a reasonable doubt as to his intent to defraud, i.e., that he held an honest belief that his conduct was not improper or unlawful, a belief the jury may have found more plausible in light of his supervisors’ approval of his colleagues’ substantially similar behavior.  

    (Citations omitted.) He also made another argument, basically that the prices were fair so it doesn't matter that he lied about them; the trial court excluded this argument, and the Second Circuit agreed. Here's how Bloomberg characterizes his defense generally:

    Litvak didn’t dispute that he made misstatements, arguing at trial that they weren’t material and he didn’t think the other parties in the transactions would be harmed because he was selling the bonds at “fully disclosed and agreed upon fair prices” that stayed below Jefferies’ threshold for 4 percent profit.

  11. And others, I guess. The one about defrauding the government, for instance. (See footnote 7.) But this is the real one.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Levine at

To contact the editor responsible for this story:
Zara Kessler at