Money Stuff: JPMorgan Undercharged for a Trade

I wrote last week that, “if you are a big bank, somebody somewhere in your organization is pretty much always doing something that looks a bit like securities fraud, and periodically the SEC will add it all up and send you an invoice for a big fine.” The next day the US Securities and Exchange Commission sent JPMorgan Chase & Co. an invoice for $151 million for assorted securities misdeeds over the last few years. I’m not sure it’s even fair to say that the SEC sent an invoice. These days it’s more like a tax return: JPMorgan downloads a form that is like “what securities laws did you violate this year,” and it fills out the form and attaches the relevant receipts and sends it all in to the SEC with a check. From the SEC’s announcement:

“JP Morgan’s conduct across multiple business lines violated various laws designed to protect investors from the risks of self-dealing and conflicts of interest,” said Sanjay Wadhwa, Acting Director of the SEC’s Division of Enforcement. “With today’s settlements, which include multiple self-reports and large voluntary payments to harmed investors, JP Morgan is being held accountable for its regulatory failures.”

The violations are very miscellaneous; there’s one about putting clients into higher-fee mutual funds rather than lower-fee but otherwise identical exchange-traded funds, and another one about waiting too long after initial public offerings to sell stocks for clients. But here’s a good one about a money market fund regulatory arbitrage. 

In March 2020, the Federal Reserve created a “Money Market Mutual Fund Liquidity Facility,” or MMLF, to help bail out money market funds. Investors had parked their cash in money market funds, but they were nervous about Covid and withdrew a lot of cash. Meeting those withdrawals would require the money market funds to sell assets, but the markets were nervous enough that there might not be any buyers. This is a pretty clear case for Fed intervention: When depositors want their money back, and banks have good assets that they temporarily can’t sell to meet those withdrawals, the central bank is supposed to step in to lend against those assets and prevent a run on the bank. So it did.

Of course money market funds are not technically banks and the Fed is not technically supposed to be a lender of last resort to them, but those feel like quaint technicalities and the Fed was not too troubled by them. As a concession to these technicalities, though, the Fed did not lend money to money market funds: Instead, it “made loans available to eligible financial institutions secured by high-quality assets purchased by the financial institution from money market mutual funds.” The money market funds had to sell their assets to a bank or broker, which would then get the (non-recourse) loan, but the promise of the Fed loan made it easier for the funds to sell without dislocating markets.

Because the Fed was looking to support US money market funds, it made only US funds eligible for the MMLF.

J.P. Morgan Investment Management runs US money market funds and also a foreign money market fund that was not eligible for the MMLF. In March 2020, its US and foreign money market funds faced redemption requests. The US funds raised money from MMLF as intended, but the foreign fund couldn’t. JPMorgan sensibly thought “well these are money market funds, it’s all short-term low-risk stuff, surely there is a trade here.” The trade is extremely simple:

1. The foreign fund sells assets to the US fund.
2. The US fund sells assets to a bank or broker eligible for the MMLF.
3. The broker brings the assets to the Fed for an MMLF loan.

All of the assets are pretty fungible, so “US money market funds can (indirectly) sell their assets to the Fed” also means “foreign money market funds can sell their assets to US funds who can sell them to the Fed.” JPMorgan’s US funds are in the business of investing in short-term cash-equivalent securities, and being in the middle of this trade is in fact a low-risk short-term cash-equivalent sort of trade. The foreign fund goes to the US fund and says “we’ll pay you a 1.3% annualized yield to hold this stuff for us overnight and then hand it to the Fed in the morning,” and the US fund is like “yeah 1.3% is a good rate for this safe trade” and does it. [1]

Is it cheating? Was JPMorgan tricking the Fed into giving liquidity to foreign funds? I mean, sure, I guess, maybe, but (1) it’s the mild loophole-based sort of cheating that you’d expect from fixed-income traders and (2) JPMorgan did ask the Fed and the Fed was like “eh fine I guess”:

On March 23, 2020, in coordination with JP Morgan IM, personnel at J.P. Morgan Securities LLC (“JP Morgan Securities”) emailed the Fed seeking guidance on whether a hypothetical transaction would comply with the intent of the MMLF. The hypothetical transaction posed to the Fed was similar to what later became the Repack ABCP trades. In the following days, the Fed and JP Morgan Securities engaged in further communications about the hypothetical transaction. The Fed ultimately responded that it did not plan to issue any guidance that would prevent the hypothetical transaction.

So JPMorgan did this to the tune of $4.3 billion of assets.

But last week the SEC objected to the prices of the trades. Everything involved here is a short-term, highly rated cash substitute, so the price of everything is more or less 100 cents on the dollar, but not quite. The foreign funds’ assets had appreciated, [2] so they had about $1.5 million of gains on $4.3 billion of assets, or about 3 basis points. There were fees to be paid to the broker that bought the assets from the foreign fund and sold them to the US funds, [3] and to the broker that bought them from the US funds to deliver into the MMLF. And then the US funds needed to be paid for buying the assets — but, again, they were buying safe low-risk assets for one night before handing them over to the Fed, so they didn’t need to be paid much. Specifically they needed to be paid a 1.3% annual yield on $4.3 billion for one day, or about $150,000.

The SEC objects that (1) the US funds took on some risk (what if the Fed had said no to taking these assets? It didn’t) and (2) they didn’t get paid enough:

The Domestic Funds earned only one-tenth of the investment gain that the Foreign Fund made from the transactions. Because JP Morgan IM personnel were involved in all aspects of the transactions, JP Morgan IM could have allocated more of the investment proceeds from the transactions to the Domestic Funds by increasing the annualized yield of the Repack ABCP but did not do so.

There was a conflict of interest here in that JPMorgan funds were on both sides of the trade, and the SEC isn’t happy with how the conflict was managed. Oh well. It is a good story of the traps of financial engineering. The US government is like “we want to backstop US money market funds,” and you are like “how can I use this to backstop my non-US money market fund,” and you come up with the sensible answer “have my US funds backstop my foreign fund and then have the Fed backstop the US ones,” and you call the Fed and say “does this work,” and they sigh and say “we can’t stop you,” and then you get in trouble with a different bit of the US government because you didn’t pay your US fund enough of a commission. 

Two ways you can invest with JPMorgan Securities are:

1. Your JPMorgan financial adviser can park your money with some outside portfolio manager, and you pay (1) that manager’s fee plus (2) a “wrap fee” to JPMorgan, or
2. Your JPMorgan financial adviser can manage your money herself, and you pay one fee to JPMorgan.

When I describe it like that presumably you immediately understand that:

1. Paying one fee to JPMorgan, instead of two fees to JPMorgan and the outside manager, is cheaper for you.
2. But paying one fee to JPMorgan, instead of two fees to JPMorgan and the outside manager, results in a bigger fee for JPMorgan. The one fee is bigger than its share of the two fees. So JPMorgan, and in particular the JPMorgan adviser who gets paid based on revenue, would prefer that you let her manage your money rather than sending a lot of revenue away to a third-party manager.

And so your JPMorgan adviser will say “you can let me manage your money, or you can use our third-party manager programs, but letting me do it is cheaper,” and then she will probably go on to say how good she is at managing money and how she won’t let you down etc. And, for a while, she would not then go on to say “but I should warn you that I have a conflict of interest: I get paid more for managing your money myself than if I find you an outside manager.” And this, says the SEC, was deceptive:

In part because clients do not pay a separate third-party fee when invested in PM Program strategies [where a JPMorgan adviser manages the portfolio], JP Morgan Securities and its financial advisors are able to charge a higher wrap fee for the PM Program, while maintaining a lower overall fee for the client. The opportunity to charge a higher wrap fee for PM Program strategies creates a financial incentive for JP Morgan Securities and its financial advisors to recommend the PM Program over the TPM Programs [where a third-party manager runs the money]. …

The PM Program brochure disclosure used until August 2021 did not disclose the conflicts of interest that PM Program financial advisors have when recommending that clients invest through the PM Program over the TPM Programs, particularly the fact that financial advisors most often negotiate a higher JP Morgan Securities fee when clients participate in the PM Program, which has no separate portfolio manager fee, instead of in TPM Programs where clients also pay a separate portfolio manager fee.

Okay. It would be weird if the JPMorgan adviser got paid the same amount for managing your money herself as she would for outsourcing it to a paid third party? But I guess you have to be really clear.

My crude rule of thumb is that if you are a giant traditional asset manager you can charge fees of roughly 10 basis points, so at 10x earnings you should be valued at roughly 1% of assets under management. And if you are a big alternative asset manager you can charge fees of roughly 100 basis points, so you should be valued at roughly 10% of assets under management. And then for some reason Pershing Square is worth 56% of its assets under management.

This suggests that if you are a giant traditional asset manager with trillions of dollars under management, and you can rebrand yourself as also a little bit of an alternative asset manager, there is a lot of valuation leverage there. If you run $11 trillion and are valued at 1% of AUM, and then people start thinking “well they’re kind of an alts manager so should be 10% of AUM,” then, uh, that adds $1 trillion to your valuation? It doesn’t really work that way, but the thought process is not entirely wrong. Here’s this:

BlackRock’s clients are pouring money into its core stock and bond offerings. To keep the momentum going, chief Larry Fink wants to push the world’s largest asset manager into the more lucrative world of private markets.

Managing more assets such as private equity, private credit, real estate and infrastructure would allow BlackRock to compete with the biggest alternative asset managers. It could also make BlackRock more valuable.

Firms such as Blackstone, Apollo Global Management and KKR manage just a fraction of BlackRock’s $11.5 trillion in assets. Yet those rivals command market values that are in the ballpark of BlackRock’s, which is around $150 billion.

The reason: Private-market funds can charge more than BlackRock gets for much of its plain-vanilla, index-based funds. And the market rewards that. ...

Alternatives made up just 3% of BlackRock’s total assets in the third quarter, but generated 11% of total revenue, highlighting how lucrative the fees are.

[Martin] Small, BlackRock’s finance chief, called out that opportunity, noting that insurance companies have $700 billion of assets with BlackRock, and flipping just 10% of that to private-credit strategies would be a huge boost to alternative assets.

The way that I tend to think about private credit is that in the olden days insurance companies bought bonds and held them to maturity, and then there was a long interlude where insurance companies invested in actively managed bond strategies, and then private credit was invented so that insurance companies could make loans and hold them to maturity. You could tell that story with a slightly different emphasis:

1. In the olden days insurance companies bought bonds and held them to maturity.
2. Asset managers realized they could charge insurance companies higher fees for actively managing their bond portfolios.
3. Over time, competition heated up and those fees compressed.
4. Asset managers realized they could charge insurance companies higher fees for private credit.

It does seem like a waste for BlackRock to charge a few basis points for bond beta when it could charge a few percentage points for “the beta of alts.”

It is roughly true that Chinese law prohibits foreign investors from owning certain Chinese technology companies. A lot of those companies want to go public offshore (in the US, UK, etc.), raise money from foreign investors and have stocks that trade freely in global markets. (Or, rather, they wanted to: This was a popular thing for Chinese companies to do for a while, but it more or less stopped in 2021.) How can you have stock that trades freely among foreigners, if you can’t be owned by foreigners? Well! There is a solution. The solution is:

1. You’ve got a Chinese company owned by Chinese people.
2. You set up another company (in the Cayman Islands or wherever) that can sell stock to anyone, and it goes public in London or New York.
3. The Chinese company and the Caymans company sign a series of contracts saying things like “the Chinese company will give the Caymans company its profits” and “the Caymans company can appoint the executives of the Chinese company,” but in vaguer ways. (“[Caymans Company] or its designated parties have the exclusive right to provide [Chinese Company] with comprehensive technical support, consulting services and other services, and [Chinese Company] agrees to pay services fees, the amount of which is determined by [Caymans Company] on the basis of the work performed and commercial value of the services” is a good way to say “the Chinese company will pay dividends to the Caymans company,” without using those words. [4] )

This is called a “variable interest entity”; we talked about it once when DiDi Global Inc. went public in New York with a VIE. So the Caymans company (which is owned by global shareholders) doesn’t own the Chinese company, in the standard legal sense of share ownership, but it has something more or less economically equivalent to ownership of the Chinese company. It quasi-owns the Chinese company: It owns the Chinese company enough to make shareholders willing to invest, but not so much that the Chinese authorities crack down.

Again, this was the state of the art in 2021, but then both Chinese and US authorities did crack down, so, uh, never mind. Still the broad idea is correct. The broad idea is that if you live in Country X and you want to own Asset Y in Country Z, and Country Z prohibits residents of Country X from owning Asset Y, you enter into some sort of agreement with someone in Country Z that gives you ownership-like rights over Asset Y without actually owning it. The concept of “ownership” is broad and fuzzy, and you can probably find something that is enough like ownership for you but not so much like ownership that the Country Z authorities will object.

Though a big chunk of the job of Country Z’s regulators is writing the rules in a way that covers all the sorts of quasi-ownership that, if they thought about it, they would object to. Still it’s probably easier for you to find loopholes than for them to close them.

Anyway here’s a story about US farmland and Chinese investors:

Walton Global has been identified by the U.S. government as a Chinese owner of U.S. farmland for a decade. The private land-banking company has opened four new offices in China since 2018 and last year was named by the U.S. Department of Agriculture as one of the top five Chinese owners of American farmland.

But last month, the company successfully petitioned the agency to reclassify much of its land as owned by investors from other countries, after The Wall Street Journal inquired about its holdings. It said the agency had made a mistake in saying so much of its land was held by Chinese investors. …

Few agree on what even counts as owned by China or which aspect of that ownership is bad for the U.S., even when that land is close to military installations. …

“We do business in China. We’re proud to do business in China,” the company’s chief executive, Bill Doherty, said in an interview. But he said, “The company is owned by me and my family. And I’m most definitely not Chinese.”

The company has touted its proximity to military installations, along with other local attractions, in some of its marketing materials in China. …

Walton said it has investors from around the world who can take brief tours of their land holdings, but don’t otherwise have access to the land, which the company then aims to sell to developers.

I don’t know enough about the ownership structures here to really comment, but “few agree on what even counts as owned by China” does seem like a pretty standard problem in these sorts of regulatory regimes.

One thing you can do is:

1. Take a job at a quantitative asset manager.
2. Email yourself all of the company’s models and secret sauce.
3. Use it to set up your own quantitative hedge fund. 

Compared to painstakingly building the models and discovering useful signals yourself, this is a pretty convenient shortcut. There are disadvantages:

1. Painstakingly doing it yourself might make you better at it: Your employer’s signals probably won’t be that good for that long, and if you just downloaded them you probably won’t understand them well enough to keep improving the models and finding new trades.
2. If you copy your employer’s models, you’ll also end up copying their trades, which might get crowded and lose their edge.
3. You’ll get, like, extremely sued? And arrested? Not legal advice, but this is very much not allowed, and people do get arrested for it.

Here is a potential solution to Problems 2 and 3:

1. Be a citizen of the People’s Republic of China.
2. Come to the US and take a job at a US quantitative asset manager.
3. Go to China and log into your work computer remotely.
4. (Use a virtual private network, because your work computer doesn’t actually allow logins from China.)
5. Download all the models and secret sauce.
6. Use them to set up your own quantitative hedge fund in China, trading Chinese stocks.
7. Get charged with a crime in the US, but don’t get extradited.

I guess? Not legal advice or anything but here’s this:

A co-founder of high-flying Chinese quantitative hedge fund Pinestone Asset Management Co. has been indicted in the US for alleged theft of trade secrets, according to people with knowledge of the matter.

Xiao Zhang, a 33-year-old Chinese citizen from Shanghai, was indicted by a federal grand jury in Boston for allegedly stealing secrets from an unidentified global investment management firm while he was working for it in 2021, according to a statement dated Oct. 31 from the US Attorney’s Office in Massachusetts. 

Here are the Justice Department press release and the indictment:

According to the indictment, in 2021, Zhang allegedly utilized a virtual private network (VPN) to access his employer’s network from the PRC, which enabled him to circumvent the company’s controls. Zhang then allegedly made copies of his employer’s code, projects and research, and sent the copies through a PRC-based file-sharing application, enabling him to again evade his employer’s controls. It is alleged that Zhang then utilized the stolen items with the intent of establishing his own investment firm in the PRC.

I will say it seems hard to run a quant fund in China, and the intuitions that you develop from long experience with finding useful equity signals in the US might not be all that valuable when applied to China. But maybe the models still are.

People are worried about the basis trade. Wall Street frenzy creates $11bn debt market for AI groups buying Nvidia chips. Nvidia Set to Replace Intel in the Dow Jones Industrial Average. B. Riley Chairman Is ‘Personally Sick’ as FRG Goes Bankrupt. TGI Fridays Files for Bankruptcy Following Years of Diners’ Declining Interest. China piles pressure on rich people and companies to cough up taxes. State Street Asks SEC for Blessing to Fit ETFs into 401(k) Plans. The SALT Deduction Fight Is Coming Back—Whoever Wins the Election. “Doctors are like any other employee, and that’s how the new generation is behaving.” Even Some High-Income Americans Can’t Afford New Cars Anymore. “When he reached the finish line 12 minutes later, ‘you could tell the dogs were really happy and excited at the accomplishment.’” Instagram Plans to Use AI to Catch Teens Lying About Age. Meta’s plan for nuclear-powered AI data centre thwarted by rare bees.

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!