Wall Street

Goldman Banker and Fed Regulator Bonded Over Steak, Leaks

When regulation is complex and subjective, regulators are especially valuable.

Where the elite meet for meat and leaks.

Photographer: Spencer Platt/Getty Images.

Banks in the U.S. are supervised by an overlapping system of regulators who conduct periodic examinations and issue numerical ratings of the banks' risk management, compliance, information technology and other areas of regulatory concern. Here's one way to find out the results of those examinations:

On September 23, 2014, the Associate attended the birthday dinner of the New York Fed Employee at Peter Luger Steakhouse, along with several other New York Fed employees. Immediately after the dinner, the Associate emailed the Managing Director, divulging confidential information concerning the Regulated Entity, specifically, the relevant component of the upcoming examination rating. The Associate wrote, "…the exit meeting is tomorrow and looks like no [change] to the [relevant] rating. I heard there won't be any split rating… [The Regulated Entity] should have listened to you with the advice…hopefully [the CEO] will now know you didn't have phony info."

That is not the official way to get the examination results -- in fact, it is distinctly frowned upon -- but honestly didn't you just assume that Peter Luger's was full of people swapping confidential information about banks? That paragraph is from this New York State Department of Financial Services action against Goldman Sachs, which agreed to pay $50 million to settle charges that, basically, it allowed this sort of thing to go on. 1  "Associate" is Rohit Bansal, then an associate in the financial institutions group at Goldman, formerly a bank examiner at the New York Fed. The "New York Fed Employee" is Jason Gross, a former co-worker's of Bansal who allegedly passed him New York Fed information after Bansal started working at Goldman. 2  The "Regulated Entity" is a "mid-sized New York bank" regulated by the DFS, the New York Fed and the Federal Deposit Insurance Corporation. Bansal had supervised the bank when he was at the Fed, and then worked on the Goldman team advising it about a potential transaction and a supervisory examination that was relevant to that transaction. That supervisory examination was reaching its conclusion just as Gross was celebrating his birthday. And what else would you talk about at a birthday dinner but the examination ratings of a mid-size New York bank? You can be absolutely certain that this was not the most boring conversation at Luger's that night.

The New York DFS, as I've mentioned before, knows how to write an enforcement press release, and the boldface bits of this one are gulp-inducing. Bansal and Gross did bad, sure:

During his employment at Goldman, the Associate wrongfully obtained confidential information, including approximately 35 documents, on approximately 20 occasions, from a former co-worker at the New York Fed (the "New York Fed Employee").

But Goldman's supervision of Bansal also looks surprisingly terrible. It's one thing to have a rogue employee secretly getting confidential information from a regulator. But Bansal seems to have been pretty open about it at Goldman:

On numerous occasions, the Associate provided this confidential information to various senior personnel at Goldman, including the Partner and the Managing Director, as well as a Vice President and another associate who perform quantitative analysis for Goldman. In several instances where the Associate forwarded confidential information to other Goldman personnel, the Associate wrote in the body of the email that the documents were highly confidential or directed the recipients, "Please don't distribute."


On August 18, 2014, the Associate shared three documents pertaining to enterprise risk management with the Managing Director, writing, "Below is the ERM request list, work program and assessment framework we used for ERM targets. Again this is highly confidential as its not public and has not been issued a[s] guidance yet. Not sure where it is at anymore due to internal politics. I worked on this framework and guidance within the context of a system working group with the Fed system. We ran several pilots to test it was well. Please don't distribute." The Managing Director replied, "I won't. Will review on plane tomorrow to DC."

Bansal sent his bosses confidential regulatory information, told them it was confidential regulatory information, and their reaction was, like, thanks, I'll take a look? 3  That's not great! Nor is the fact that when Bansal left the New York Fed, 4  he was legally restricted from working on behalf of the bank, he told Goldman about that restriction, but Goldman nonetheless "placed the Associate on Regulated Entity matters from the outset of his employment." 5  

Of course it did! "The Associate was hired in large part for the regulatory experience and knowledge he had gained while working at the New York Fed," says DFS. "The leak has intensified scrutiny of both the so-called revolving door -- employees moving between regulators and banks -- and the specific relationship between Goldman Sachs and the New York Fed," says Bloomberg. "While the so-called revolving door is common on Wall Street, the investigation into the Goldman banker and the New York Fed regulator exposes the perils of the job hopping, affirming the public’s concerns that regulators and bankers, when intermingled, occasionally form unholy alliances," says the New York Times

But ... how unholy? The DFS consent order is intentionally vague on details, but it is hard to tell what actual bad results occurred here. Goldman at least had the opportunity to use Bansal's information in sort of a gross ambulance-chase-y way to pursue business based on its uncanny insights into the regulatory mind. There's that line above about how "hopefully [the CEO] will now know you didn't have phony info," and at one point Bansal suggested pitching stress test advice to a bank that "was soon to be subject to a nonpublic enforcement action." It's not clear that that pitch actually happened, though, or that Goldman ever used any of this information for, as it were, marketing.

Mostly, what seems to have happened is that Bansal used his experience, and his illegally obtained documents, to help Goldman give its clients better regulatory advice. So he shared that enterprise risk management assessment framework, which had "not been issued a[s] guidance yet." (But was supposed to be eventually?) He gave this fairly bland advice:

The Associate also provided advice to relay to the Regulated Entity’s management, stating that they should “keep their cool, not get defensive and not say too much unless the regulators have a blatant fact wrong” as it “will go off better for them in the long run. Believe it or not the regulator’s [sic] look for reaction and level of mgmt respectiveness [sic] during these exit meetings.”


In another instance, on August 18, 2014, the Associate emailed the Vice President and the other associate photographs of a projected PowerPoint presentation on stress testing, which he had made to the Regulated Entity while employed at the New York Fed, writing, “PLEASE do not distribute as this is confidential.”

And the misdeed that ultimately got him caught:

On September 26, 2014, Goldman had an internal call regarding the calculation of certain asset ratios, during which there was disagreement over the appropriate method. During the call, the Associate circulated an internal New York Fed document – which the Associate had recently obtained from the New York Fed Employee – relating to the calculation, to the call participants, writing, “Pls keep confidential?”

So Goldman was wondering how its bank client should think about enterprise risk management, or prepare for a stress test, or show deference to regulators in a meeting, or calculate some ratios. And Bansal -- the guy Goldman had hired from the New York Fed for his "regulatory experience and knowledge" -- had a pretty good idea of what the New York Fed thought about enterprise risk management, and stress tests, and deference, and ratios. So he told his Goldman co-workers, presumably so they could tell the client.

Is that bad? It seems like cheating, in the sense of having the answers to the test before taking it. But ... don't we want the bank to get the answers right? Like, if the bank was calculating its ratios wrong, and Bansal knew how to calculate them right, shouldn't he have told the bank? Wouldn't that make the bank safer? And if the New York Fed knows how banks are supposed to calculate their ratios, and keeps that secret from the banks, isn't that ... weird?

Not entirely, of course: With the stress tests, for instance, the Fed has been clear about the need for banks to do their own work and come to their own conclusions, rather than just guessing at what the Fed will want to hear. That's a good way both to make sure that the banks are being thoughtful about risk management, and to avoid a monoculture of thinking about risk that might lead to systemic dangers. There are genuinely good reasons not to make compliance too easy.

But mostly when I think about the regulatory revolving door, I don't think about it as providing incentives for regulators to be too lax and deferential with banks, so that the banks will love them and hire them at huge salaries. Instead, I think of it as providing incentives for regulators to make regulations complex, opaque and subjective, so that the regulators' own expertise becomes extra valuable in the private sector. That seems to have been Bansal's main value to Goldman: There were lots of gray areas of bank regulation that were subject to interpretation, and he was able to help Goldman figure out what the New York Fed's interpretation was likely to be. What got him in trouble is that he did that mostly by circulating New York Fed documents that he wasn't supposed to have, which seems like a rookie mistake.

Because bank regulation is itself so complex and uncertain, it requires regulated banks to understand not just the letter of the law but also the minds of the regulators. One way to try to understand the minds of regulators is to read their internal guidance and presentations, but that is wildly illegal, and puts you in line for a $50 million fine. Just hiring former regulators, on the other hand -- and not letting them obtain lots of secret documents from their former colleagues -- remains attractive. If bank regulation is largely about getting into the minds of regulators, then the minds of regulators will always be valuable on the open market.

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

  1. In addition to the $50 million, Goldman "accepted a three-year ban on some advisory work in New York," and "Bansal and Gross are expected to plead guilty to misdemeanor counts brought by the Manhattan U.S. Attorney."

    Oh, also, disclosure: I used to work at Goldman Sachs, though I never went to Peter Luger's and pried confidential information out of regulators.

  2. "Who has since been terminated for this conduct," obviously. (So has Bansal, from Goldman.)

  3. Eventually Bansal's ultimate boss, a Goldman partner, reported him to compliance:

    On September 26, 2014, Goldman had an internal call regarding the calculation of certain asset ratios, during which there was disagreement over the appropriate method. During the call, the Associate circulated an internal New York Fed document – which the Associate had recently obtained from the New York Fed Employee – relating to the calculation, to the call participants, writing, "Pls keep confidential?" Following the group call, the Partner called the Associate to discuss the document, including where he had obtained it, and the Associate told him that he had obtained it from the New York Fed. The Partner then called the Global Head of IBD Compliance to report the matter and forwarded the document.

  4. By the way, DFS says that he "was required to resign from his position at the New York Fed for, among other reasons, taking his work blackberry overseas without obtaining prior authorization to do so and for attempting to falsify records to make it look like he had obtained such authorization, and for engaging in unauthorized communications with the Federal Reserve Board." Which is ... also nuts? Like I am not that bothered by the Blackberry thing, but the falsifying records thing is troubling, and what does "unauthorized communications with the Federal Reserve Board" even mean? Was he, like, pestering Daniel Tarullo about how that mid-size New York bank should be regulated?

  5. It's possible that there's some interpretive wiggle room on this question:

    Accordingly, the Associate inquired with the New York Fed Ethics Office and was given a "Notice of Post-Employment Restriction," which he completed and signed with respect to his supervisory work for the Regulated Entity. The Associate provided this form to Goldman. This Notice of Post-Employment Restriction read that the Associate was prohibited "from knowingly accepting compensation as an employee, officer, director, or consultant from [the Regulated Entity]" until February 1, 2015.

    On May 14, 2014, the Associate forwarded this notice of restriction to the Partner, the Managing Director, and an attorney in Goldman's Legal Department. In his email, the Associate also included guidance from the New York Fed, stating, in short, that a person falls under the post-employment restriction if that person "directly works on matters for, or on behalf of," the relevant financial institution.

    Despite receiving this notice and guidance, Goldman placed the Associate on Regulated Entity matters from the outset of his employment. 

    Those terms are a bit vague, and I guess it's possible that just providing support to the Goldman bankers who advised the bank -- as opposed to, say, dealing directly with the bank, or lobbying the New York Fed on the bank's behalf -- could have been viewed as permissible. 

To contact the author of this story:
Matt Levine at mlevine51@bloomberg.net

To contact the editor responsible for this story:
Zara Kessler at zkessler@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.