Money Stuff: Please, Mr. Potato Chip Is My Son’s Name

BloombergOpinion Money Stuff

Matt Levine

Mr. Potato Chip As I mentioned yesterday, the U.S. Securities and Exchange Commission has been bringing a deluge of enforcement actions recently to close out its fiscal year, and all I want to do is write about them. So let's do that. Here  are some things that the SEC has to say about James Wallace Nall, III: Over the last 15 years, Nall has developed close personal and business relationships with a group of individuals residing in or near Thomaston, Alabama, a rural, 400-person, "one-traffic-light" town located in southwestern Alabama. ... Nall's public persona in Thomaston was closely associated with Golden Enterprises. Nall openly discussed, and all of the defendants were aware of, his father's role with the company. More generally, Nall was known throughout Thomaston as "Mr. Potato Chip." Here is one thing that the SEC has to say about Nall's father, whose name is James Wallace Nall, IV, in what is perhaps my favorite footnote in the history of SEC enforcement actions: It is unusual in this matter that Nall III is the son, while Nall IV is his elder father. It is unusual! You probably know where this is going; when the SEC writes Southern Gothic it always ends up being about insider trading. I kind of wish it wasn't. I kind of wish that the SEC enforcement lawyers, confronted with the "Mr. Potato Chip" of small-town Alabama and his possibly time-traveling father, had just abandoned their insider trading case and moved to Thomaston to record a longform podcast series examining the town's quirky characters and dark secrets.  But, no, insider trading. Golden Enterprises Inc. was a publicly traded potato chip manufacturer, which you might know from its Golden Flake brand of chips. In 2016 it was acquired by Utz Quality Foods LLC, another potato chip manufacturer, which you might know from its Utz brand of chips. Nall IV sat on the board of directors of Golden Enterprises and learned of the merger well before it was announced. He told Nall III, for good and legitimate reasons: Nalls III and IV had a family business that owned Golden Enterprises shares, and Nall IV wanted to discuss the potential tax consequences with his partners, including Nall III. He "specifically informed the three men that the information was highly confidential, and that they should not engage in any trading in Golden Enterprises stock," and "further admonished that 'nobody that we know can buy any stock, period.'" And then Nall III, who had a reputation as "Mr. Potato Chip" to live up to, allegedly went and told some of his friends, who traded on his tips and made money when the merger was announced. Neither Nall traded for their own accounts. The SEC sued Nall III and his friends who traded.  Notice that Nall IV told Nall III about the merger and didn't trade, and Nall III allegedly told his friends about the merger and didn't trade, but the SEC is going after III and not IV, because IV had a good reason for telling III and swore him to secrecy, while III had no good reason for telling his friends and allegedly expected them to trade. All of insider trading law turns on subtle distinctions like this: If you tell someone about a merger and expect them to keep it secret you're fine; if you tell them and expect them to trade then you're in trouble. This makes sense, sure, but it can be a little hard to prove one way or the other years after the fact. Connoisseurs of insider trading law know that it is also generally important for the authorities to prove that you received some "personal benefit" for tipping the people who traded. This has its own legal rationale, but it is also very useful as a way to demonstrate your intent. If you tell someone about a merger, and they trade, and then they hand you a paper bag full of cash as your cut of the profits, it is hard for you to argue that you didn't expect them to trade. Let's see what personal benefit Nall III allegedly got for his tips: Shortly after Nall learned of the proposed sale of Golden Enterprises (and tipped Hale Smith), Nall and Hale Smith developed a plan to help revitalize Thomaston's "dying" downtown by purchasing and renovating an old "general store" and potentially using it as a deer-hunting museum. Immediately before Tutt began trading in Golden Enterprises stock, he agreed to relocate his real estate business to the top floor of that building. See, the case will turn on whether Hale Smith agreed to support a deer-hunting museum out of genuine civic pride or as a disguised kickback for merger tips. Insider trading law is amazing. The alleged tippees all settled without admitting or denying guilt. Nall IV, who allegedly told Nall III about the merger in confidence, is not accused of any wrongdoing. Nall III has not settled, and the SEC does not seem to have much in the way of smoking-gun evidence against him. If this case goes to trial I … might have to go watch? I might have to make this podcast myself, and the trial is as good a place as any to start. Audit conflicts The way auditing works is that a company does a bunch of work to put together its financial statements, and then it shows the financial statements to an outside accounting firm, and the outside accountants say "yeah these statements are fine, we will sign our name to a statement saying that they look fine." And then everyone tends to trust the financial statements because an the auditors have signed off on them.  One weird thing about this is that sometimes the company will hand the auditors some financial accounts and the auditors will say "no you are not supposed to do these accounts this way" and the company will say "ah fair enough, you are expert professionals and we are not, we have messed up a fine point of bookkeeping, glad we hired you to check." There will be a pause, and then the company will say "well but how are we supposed to do these accounts?" And the auditors will say "we can't tell you that! We are auditors, and our sacred role is to check your accounts for you; if we just did the accounts for you then we would be checking our own work and there would be terrible and irreconcilable conflicts of interest. You will have to figure it out for yourself sorry." I am exaggerating a little. The auditors can give hints. It usually works out fine. But they can't do the work for the client; the job of keeping the books and preparing the financial statements has to be kept separate from the job of auditing the financial statements.[1] It can be annoying for the client! You are paying these people a lot of money, and in exchange they will tell you that your accounts are wrong, but they will not make them right. Here's a funny Securities and Exchange Commission enforcement action against PricewaterhouseCoopers LLP and a PwC partner named Brandon Sprankle. PwC was auditing a public company that the SEC just refers to as "Issuer A." As part of the audit Issuer A had to deliver reams of stuff to PwC, and it apparently needed to build a doodad called a "GRC" to generate some of that stuff. It is tempting to just leave it at that; you will not understand the situation any better if you know what "GRC" stands for or what it did. (Nonetheless.[2]) If Issuer A couldn't deliver the output of the GRC to PwC, PwC would make disappointed noises, so Issuer A needed to build the GRC to make PwC happy and complete its audit. PwC knows how GRCs work and how to build them, and Issuer A did not. So Issuer A quite sensibly asked PwC, well, can you just build this GRC thingy for us. And Sprankle, an auditor on the account,[3] committed a grave auditing sin: He built the GRC for them. Helpful! Not allowed![4]  The bulk of the SEC's order is about how Sprankle told Issuer A that PwC would build the doodad for them, while telling his internal conflicts-of-interest-review team that he would only audit the doodad: In seeking internal authorization to perform the non-audit GRC work, Sprankle drafted an engagement letter for approval by PwC's Risk Assurance Independence ("RAI") group, an independence-reviewer within his business unit. In the draft engagement letter, Sprankle described the proposed services as assessing multiple areas, and providing observations and recommendations, as opposed to designing and implementing the GRC project. … In early June 2014, Issuer A again put PwC on notice that it expected PwC to design and implement a GRC solution for Issuer A and to manage the project. After PwC sent the draft engagement letter, Issuer A's then-Head of Internal Audit objected to the description of the services contained in the draft engagement letter. In the email, he informed PwC that the proposed work was an "implementation project that's been outsourced" to PwC. ... From August through mid-October 2014, PwC managed the project, performed substantial design work, configured the design on a non-production server, and provided oversight and direction for the implementation to a live environment. According to the senior manager for IT Internal Audit: Issuer A had little involvement in the assessment and design phase of the project; further, Issuer A lacked the technical expertise to configure the system; and, although Issuer A ultimately had to approve the work, PwC exercised decision-making authority in designing and configuring the GRC module. See, "design," "implement," "manage," these are all bad words; auditors can't do those things for the companies they audit. "Assessing" and "providing observations and recommendations" are fine. You can give the company some hints about what the right answer is, but you can't tell them the right answer. This is all perfectly sensible, by the way; I sympathize with the client who is like "if you think we need a GRC just build a GRC for us," but I can understand why the SEC forbids it. For one thing, there is the obvious financial conflict: If the auditor is getting paid to build software and do other non-audit stuff, it will be more inclined to let some problems slide in the audit to keep up the relationship. (In settling with the SEC PwC agreed to pay about $7.9 million, including disgorging about $3.8 million of fees that it collected for non-audit projects.)  But what is interesting about these rules is that they acknowledge that there is a more important conflict than just money. The reason auditors can't prepare the financial statements and then check them is not just that they'd get paid more for doing it; the main reason is that it is an intellectual conflict of interest. If the auditors do the books then they will be proud of the books and won't want to find errors in them; if they design the company's internal controls then of course they will judge those controls to be good. The way to make an audit effective is to keep the auditors at arm's length and kind of suspicious, rather than letting them get too comfortable with the company. The auditor is a backstop to the accuracy of a company's financial statements; if it builds the software and checks its results, then you have lost one layer of checks. Anonymous bonds Basically my analysis of dark-pool scandals goes like this: People want to trade in a safe space where they can remain anonymous and where their orders are not exposed to evil high-frequency traders, so they like dark pools that promise secrecy and privacy. If the dark pools are too good at providing secrecy and privacy, though, no one will trade. If your order isn't exposed to anyone, then no one will be able to "front-run" you, but no one will be able to trade with you either. One simple way to resolve this conflict is for the dark pool to promise you lots of privacy and exclusivity, and then secretly invite in a bunch of high-frequency traders to actually execute your orders. That is bad and the dark pool will get in trouble for it. There has been a string of cases.  That is my story about equity dark pools that trade stocks. But this week the SEC brought a case against a corporate bond trading platform run by TMC Bonds LLC, where the issue was different but analogous. Here the concern is not really about evil high-frequency traders front-running orders; it's about simple anonymity. TMC promised investors anonymity: If you put in an order to buy or sell bonds on its trading system, other customers would see your order, but they wouldn't know whose order it was. This is standard in many order-driven trading systems (stock exchanges, for instance); it is also a pretty basic expectation of most bond trading. If Pimco calls a bank and says "hey can you buy a bunch of bonds for me," and the bank calls up its other customers and says "hey Pimco is looking for some bonds do you have any," Pimco is going to be disappointed. "TMC Bonds publicly touted its anonymous trading platform," says the SEC, "but, in fact, disclosed the identities of certain firms seeking to trade corporate bonds to potential counterparties over 2,500 times during a two-and-a-half year period." Sounds bad! But here is the explanation: TMC's corporate bond trading desk employees disclosed to ATS subscribers the identity of other subscribers attempting to trade on the ATS. Specifically, between January 2016 and June 2018, TMC employees on the corporate bond trading desk disclosed the identity of certain firms attempting to trade on the ATS approximately 2600 times. The subscribers whose identities were disclosed were large broker-dealers that typically traded corporate bonds on the ATS on behalf of retail customers. In most instances, TMC disclosed the name of the retail brokerage firm in situations where the firm had previously attempted to transact with an order posted on the ATS by another subscriber, but the posting subscriber failed to confirm the transaction in a timely manner. TMC attempted to facilitate an execution by disclosing the retail brokerage firm's name to the subscriber that had not confirmed the trade request in order to encourage that subscriber to proceed with the trade. Other situations in which TMC disclosed retail brokerage firms' identities included requests from those firms to cancel trades and for assistance in completing a trade, and gratuitous disclosures designed to encourage proprietary firms to trade with retail brokerage firms. Disclosures of a retail brokerage firm's name usually resulted in the trade being completed at the quantity and price sought by the retail brokerage firm, largely because TMC knew that other ATS subscribers, such as proprietary trading firms, were generally more willing to trade with retail brokerage firms than other professional traders. Proprietary trading firms generally felt that retail brokerage firms were only seeking to execute their customers' orders in a timely fashion, whereas other professional traders might have been trading to gain an informational advantage that could be used on other trading venues. TMC's disclosure of the retail brokerage firms' names helped TMC to meet the high fill rates required by those firms, and thus, keep from potentially losing those firms as subscribers. Continued retail firm participation on TMC's platform was important to the business.  A proprietary trading firm (think, roughly, "evil high-frequency trader") would post a non-binding indication of interest to sell bonds on TMC's platform, and a retail customer of a big brokerage would want to buy those bonds, so the retail broker would submit a trade request on TMC to buy the bonds. The prop firm would decline to trade for some reason. I don't know why, but as a very general statement of market structure, any time you have posted an order and someone tries to trade with it, there is a good chance you have made a mistake. Like, if you are a market maker and you offer to buy a bond for $100 or sell it for $101, you generally have no particular information about which way prices are going to move in the future; you're just putting some spread around the current price. If someone comes in and agrees to buy at $101, there's a good chance that they know something you don't, and that the right price is $102. This is called "adverse selection." This doesn't mean that you should always back away from your orders—then you'd never trade—but you will be a little nervous every time you trade, and if you are allowed to back away from your orders sometimes you will. But if you find out that the order comes from a retail customer then, fine, whatever. A retail customer presumably has no inside information about bond prices, and her order is unlikely to be the tip of the iceberg of a much larger order that will move those prices. Retail orders are assumed to be random, so you can trade with them without worrying too much about adverse selection. And so a major focus of a lot of market making is on segmenting retail orders (safe, lucrative) from institutional orders (scary, full of adverse selection). In equities this usually involves "payment for order flow," in which market makers pay retail brokers to send orders directly to them rather than going to stock exchanges; this is vastly, confusingly controversial. In TMC Bonds it involved a simpler system, which is just, the market makers would back away from orders because they were worried about adverse selection, and then TMC would send them a message saying "no no no this is retail, it's fine, you should do this trade," and they would. The SEC doesn't even suggest that the retail customers or brokerages were harmed by this. The retail brokers got what they really wanted (trades at good prices) and were told that they were getting what they thought they wanted (anonymity). If anyone was harmed it was the other customers: If a market maker backed away from an order, and didn't get a message from TMC saying "no this is retail it's fine," then that was a good hint that it was a professional order. If you violate anonymity by disclosing the names of the people whose names help them, then you are sort of implicitly violating anonymity for everyone. Treasury prices Here is a fun and sort of philosophical story about measurement. Specifically it is about how there is no official closing price for U.S. Treasury securities, so "money managers sometimes value identical securities at different levels." How different? A really very very tiny amount of different: Consider the Treasury note issued in May 2017 with a 2% coupon and maturing in 2024. There's nearly $77 billion worth of them, making it one of the largest Treasury issues in existence. Three large bond mutual funds that held it valued the debt slightly differently at the end of the second quarter. The Bond Fund of America went with 101.1880. The Federated Total Return Bond Fund used 101.1820. Northern Fixed Income Fund marked it at 101.1758.  That's a 0.0122% spread between the highest and lowest valuations, which is very low but not quite zero. Zero is, however, achievable, and is regularly achieved in other markets: By contrast, large equity funds managed by those three companies all agreed on where Microsoft Corp.'s stock ended the month of June: $133.96, which was the industrywide closing price. There are official closing prices for stocks, and so at the end of the month everyone marks their stocks to those prices. But the reason there is no variance in the official Microsoft stock price is not that there is more agreement on how much that stock is worth; it's that there is perfect agreement on the convention for picking its month-end value. There is a single trade—the closing auction in the stock's primary exchange—that counts as the official final trade of that stock, and everyone agrees to use that convention.  If there were no convention and you just looked at raw data, it might not be obvious to you what price to pick. In the last two minutes of trading this June—from 3:58 to 4:00 p.m. on June 28—Microsoft stock traded at prices ranging from $133.74 to $134.19, a range of about 0.34%, or 27 times the range of "closing" prices on that Treasury note. Meanwhile during the whole last trading day of June, that Treasury traded in a range from 101.05459 to 101.21875, or about 0.16%. There was far more widespread and stable agreement on the value of that Treasury bond on June 28 than there was on the value of Microsoft. But the stock has a convention, so everyone can go home saying, well, at 3:58 the price of Microsoft stock was a fluctuating uncertain thing, but then at 4:00 we all got a number we could write down and call it a day. And then it starts fluctuating again in after-hours trading. Anyway Tradeweb Markets Inc. and the Intercontinental Exchange Inc. are teaming up to change that by publishing closing prices for Treasuries, and, good! Obviously it is useful to have one agreed-on price at the end of the day; everyone can mark their books to it, derivatives can settle based on it, etc. It is useful to have conventions for these sorts of things, but you can't confuse the conventions with reality. Things happen KKR Has Quietly Built an Investment-Banking Contender. Masayoshi Son's unyielding optimism — and lack of challengers — led SoftBank to overvalue WeWork, sources say. SoftBank in talks to boost $1.5bn WeWork pledge. WeWork Puts Three Businesses Up for Sale. Fed Bows to Market Demand for Repo by Boosting Size of Actions. Big Banks Loom Over Fed Repo Efforts. Reforms Have Made Banks Safer but Markets More Brittle. Brexit prompts ESM to switch out of English bond law. Bank of America CEO Says His Company Has to Make Profit and 'Take a Stand.' Elliott Management Renews Push to Split Up Marathon Petroleum. Credit Suisse Tabloid Scandal Explodes Into Threat to CEO Thiam. Meet the Monster Energy Elite: The Connoisseurs Collecting the Most Extreme Drinks on the Planet. Labradoodle creator says it's his 'life's regret.'  If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks! [1] Rule 2-01 of the Securities and Exchange Commission's Regulation S-X sets out some of the rules. "An accountant is not independent," says Rule 2-01(c)(4), if it provides "bookkeeping or other services related to the accounting records or financial statements of the audit client," "financial information systems design and implementation," "appraisal or valuation services," or other sorts of accounting work that might "be subject to audit procedures during an audit of the audit client's financial statements." If the auditor does the financial statements, it can't audit them. [2] It stands for "Governance Risk and Compliance" software. "GRC systems are used by companies to coordinate and to monitor controls over financial reporting, including employee access to critical financial functions. Issuer A intended to use the GRC software to generate information as part of the company's control environment and to provide data to assist personnel in forming conclusions regarding the effectiveness of internal controls related to financial information systems. As such, at the time the GRC system was being implemented, it was intended to be subject to the internal control over financial reporting audit procedures." Auditors have to opine on the adequacy of internal controls over financial reporting, so the GRC is, in effect, an input to that audit process.  [3] Not the lead audit partner, by the way. He "participated in PwC's audit of Issuer A as an information technology specialist on the audit engagement team." That seems a bit more forgivable to me. If the lead audit partner is selling non-audit services that's bad. If the tech specialist on the audit is like "oh by the way have you tried our tech," I am more inclined to allow it. [4] As the SEC puts it: "The Commission's independence rules prohibit independent auditors from designing and implementing systems such as GRC where the software aggregates source data, or generates information significant to the clients' financial statements or other financial systems as a whole. Designing, implementing, or operating systems affecting the financial statements may also place the accountant in a management role, or result in the accountant auditing his or her own work or attesting to the effectiveness of internal control systems designed or implemented by that accountant."

Bloomberg L.P. ● 731 Lexington Avenue, New York, NY 10022 Web ● Facebook ● Twitter Feedback ● Unsubscribe