Businessman using train on phone
Eurostar

If you are an investment banker, and you settle into your seat on a train after a long day of doing investment banking, and after a while you look up from your novel and notice that the person sitting next to you is also an investment banker, a competitor, and not only that but he is doing investment banking right now, and not only that but you can see his email and figure out what company he is working for and what deal he is working on, what should you do?

There is a correct answer, and you are not going to like it, but it is: You should get out your phone and, carefully and casually and without letting him see your screen, you should email the people back at your office and try to get them to pitch his client right now and take the deal away from him, or at least get a role in it. Then your colleagues will scramble to use the information you gather to put together, within the hour, a compelling pitch to the client, and they will call the client and present their credentials and make their pitch, and the client will say "wait how did you even know we were doing a deal, it is secret," and your colleagues will smugly reply "we are investment bankers, it is our business to know what deals are going on, and we are very good at our business, also by the way if I were you I'd have some questions for your current bankers about leaks," and the client will be impressed and hire your bank and fire the other guy's bank.

Ideally all of this will happen while you are still on the train, and the client will call the other guy and be like "never mind, you're fired," and he'll slump in his seat looking stricken and confused, and then your phone will ring and you'll loudly answer it "why hello Mr. Client, yes it is a pleasure to be working with you on this deal," and he will turn to you with slowly dawning understanding, and then that's it, you win, you have done the very best investment banking thing ever, no one will ever top this story, you have completed the final level and passed into investment banking Nirvana.

This is not quite what happened to Alexandre Zaluski on a July 15, 2014, Eurostar train from London to Paris, but, man, he gave it his best shot:
The backdrop for the case is the audacious attempted takeover of T-Mobile US Inc. by the French mobile-phone carrier founded by Xavier Niel. [Lazard Ltd. banker Vincent] Le Stradic was and still is one of the French billionaire's closest financial advisers, while the man in the seat next door, Alexandre Zaluski, worked in the investment banking arm of UBS in London, focusing on equity capital markets.

During the journey, Zaluski caught a few glimpses of messages that popped up as Le Stradic juggled between phones. Unable to make sense of the piecemeal information into the deal, which would be announced two weeks later, the UBS banker decided immediately to reach out to a colleague via email.

"Niel would become CEO of T-Mobile," Zaluski told Christian Lesueur, head of telecom, media and technology for Europe, Middle East and Africa at UBS. Later on, Zaluski would also write that the messages he saw on his neightbor's phone mentioned a $40 billion deal and suggested Niel had "a vision" for the U.S.

It didn't take Lesueur long to understand what was afoot: Iliad SA was planning to force its way into the U.S. market by taking a majority stake in a wireless provider much larger than the French company.

Quickly, Lesueur put together a team of a dozen UBS bankers to make a financing pitch to Iliad.
Yes! That is how you do it. It did not work, but this sort of thing is always a long shot: "In my mind, the transaction was stillborn. There was no doubt about that," Lesueur said. "I just did my job."

It is the prime directive of investment banking, more important even than making money: If there is a deal in your sector, you had better either be involved in the deal, or have a really good answer when your boss asks you "why aren't we in this deal?" Lesueur found out that a deal was going on without him, and he clawed desperately to get into it. That really is his job!

I am recounting all of this because, bizarrely, investigators at the French Autorité des Marchés Financiers have claimed that it was illegal, and want to fine Zaluski, Lesueur and UBS for doing it:
None of the trio made any illicit trades based on the nonpublic information, but AMF officials accused Zaluski and Lesueur of transmitting the insider tip to colleagues at UBS when they shouldn't have. The crux of the argument brought forward by investigators is that since Zaluski found out about the takeover bid outside of normal work channels he wasn't allowed to share it.
Huh! Under U.S. law I think it would probably even have been legal for them to trade on information overheard on trains, but of course they didn't do that. They didn't use this information for a quick personal profit; they used it to do their jobs, in a way that was absolutely expected of them and frankly rather heroic. Zaluski "denies any wrongdoing, saying Le Stradic blamed him for his own failure to guard the secrets, while Lesueur said he was doing what he would be expected to do: collect information in order to drum up new business."

Perhaps the law in France really is that, if you are an investment banker on a train and you notice your seatmate working on a deal, you should go back to your novel and forget about it. But I am not sure how that could work, since it so clearly conflicts with the most basic natural instincts of investment bankers.

Heartbeats

The way a mutual fund works is that a bunch of people put money in a pot and the manager of the pot uses the money to buy some stocks. Sometimes the fund will also have to sell some stocks, either because one of the investors in the pot wants to take money out (and so the manager needs to sell stocks to get the money to give back to the investor), or just because the manager wants to sell some stocks to buy some different stocks. If the fund sells a stock for more than it paid for it, it has a taxable profit, which is passed on to the investors in the fund. 1 So if you invest in a mutual fund, you will sometimes have to pay taxes on the fund's trading activities.

The way an exchange-traded fund works is different. The simple way to think about the difference is: ETFs do not sell stock. 2 For one thing, lots of ETFs have traditionally been passive index funds, which just buy all the stocks in the index and hold them; their managers never decide to buy different stocks, so they never have to sell old shares to buy new ones. For another thing, when investors want to take their money out of an ETF, they do not actually "take their money out"; instead, they just sell the ETF on the stock exchange to another investor. (Thus: "exchange-traded fund.") This means that investors can get their money back without redeeming out of the fund, which means that the ETF doesn't need to sell shares to redeem investors. Since ETFs don't have to sell shares, they don't have taxable profits to pass on to their investors. And so if you invest in an ETF, you won't have to pay taxes on the ETF's trading activities (because there are none). (Technically you have only deferred taxes-when you ultimately sell your ETF shares, you will be taxed on any appreciation-but that can be a long time, and it's a nice perk.) And this is a principal selling point of ETFs. "ETFs have lower capital gains and they are payable only upon sales of the ETF," says Fidelity. "ETFs are typically structured with the aim to shield investors from capital gains taxes," says BlackRock.

That was the basic version of the story. Here's the intermediate version.

The thing I said about investors getting their money back without redeeming isn't 100 percent true. Often investors can get out of an ETF without redeeming, just by selling their ETF shares on the stock exchange, but sometimes there are more sellers than buyers and the ETF has to shrink. But the way that happens is not by investors going to the ETF and asking for their money back. Instead it is by investors-really "authorized participants," a limited group of big banks and market makers that deal directly with the ETF-going to the ETF and asking for their stocks back. The authorized participant gives the ETF a big block of ETF shares, and in exchange the ETF gives the authorized participant back a big pile of all the different stocks in the ETF. (The AP can then go sell those stocks to get its money back. 3 )

The ETF does it this way, rather than just selling the stocks itself and giving the AP the money back, partly due to the arbitrage mechanism that keeps ETF prices in line with their net asset values. But it's also about taxes. The tax rule is that if the ETF sells the stock for more than it paid for it, then that is a tax event. But if the ETF gives the stock to investors as an in-kind redemption, then that is not a tax event. This is not a fundamental necessity of tax theory-you could easily tax the ETF for redeeming with appreciated stock-but it just happens to be the rule. 4 And so, again: ETFs do not sell stock, even to meet redemptions, and so they do not create taxable income for their investors.

Okay now here's the advanced version.

The thing I said about ETFs being index funds that never need to sell shares is also not 100 percent true. Many traditional passive ETFs track broad market indexes that rarely change, and so they rarely need to sell stocks, but every so often a stock will be taken out of the index and they will need to sell it. And many other ETFs are much more active, tracking specialized (factor, proprietary, etc.) indexes that change periodically, so they sometimes have to sell large amounts of stock. But I just said that the main idea of ETFs is that they do not sell stock. So how does that work?

Well here are Zachary Mider, Rachel Evans, Carolina Wilson and Christopher Cannon in Bloomberg Businessweek with the explanation. 5 They call it a "heartbeat":
Most heartbeats take place when an ETF has to get rid of stock because of a change in the index it tracks. That was the case in September at State Street's $23 billion Technology Sector Select SPDR. Some of the fund's largest holdings, including Facebook Inc. and Google parent Alphabet Inc., were being dropped from the portfolio because they were leaving the index the fund is designed to mimic. Both stocks had more than doubled since the fund first acquired them, so it probably faced a hefty tax bill if it sold them.

On Wednesday, Sept. 19, two days before the index change, $3.3 billion of new assets poured in, increasing the fund's size by 14 percent. The additions came in the form of stock that matched the fund's holdings at the time. The investor's identity wasn't disclosed, but market participants say that, given the size of the transaction, it was probably a large investment bank.

On Friday of that week, someone pulled more than $3 billion back out. Trading data suggest it was the same investor that had appeared two days before. Rather than take back the same shares of stock it had contributed 48 hours earlier, the investor appears to have walked away with the fund's oldest shares of Facebook and Alphabet, the ones with the biggest capital gains. Thanks to winnings on stocks like that, the fund reported more than $4 billion of capital gains for the year. But since it used the ETF tax loophole to shield those gains from taxes, its annual report shows, it ended up reporting a $309 million capital loss to the IRS.
Get it? To get rid of appreciated stocks-without selling them-the ETF will give them to a big investment bank in redemption for shares of the ETF. Usually the way in-kind redemption of ETF shares works is that the bank gives the ETF a bunch of ETF shares, and the ETF gives the bank back a complete representative sample of the shares in the ETF: The redemption turns the ETF shares into the underlying portfolio. But that is not a legal requirement, and the ETF and the bank can negotiate for some other redemption basket of stock. If the ETF owns some stocks that it wants to get rid of, it can give the bank only those stocks in redemption for its ETF shares. Instead of selling them and paying taxes, it hands them out to a shareholder as part of an in-kind redemption, and avoids taxes.

But this only works if the bank actually owns a lot of shares of the ETF, so that it can redeem them. That will not generally be the case. And so to make this work, the bank will buy a bunch of shares in the ETF-or rather, do a "create" transaction by handing the ETF a basket of the underlying stocks and getting back shares of the ETF-a few days before it needs to do the redemption. 6 The bank pumps stock-sometimes billions of dollars' worth-into the ETF, only so that it can suck billions of dollars' worth of different stock out of the ETF a couple of days later. Thus, "heartbeat."

Is this bad? Obviously it avoids taxes, and it looks a little, you know ... I mean, the headline of that Bloomberg Businessweek article is "The ETF Tax Dodge Is Wall Street's 'Dirty Little Secret.'" "It's a transaction that looks too good to be true," says a tax expert. "'If the IRS were looking at it, they would say that's a sham transaction,' says Peter Kraus, a former chief executive officer of mutual fund manager AllianceBernstein Holding LP and a critic of ETFs."

On the other hand I am not convinced that one should look at the transaction in isolation here. My view of the situation is not only that "an ETF is a mutual fund that doesn't pay taxes," but also that everyone accepts that. There just seems to be broad agreement among investors and regulators and policymakers that an ETF is supposed to be tax-efficient, that ETF investors get to defer capital gains until they sell their shares. (Again: This is a very widely advertised benefit of ETFs. 7 ) Some ETFs ran into a bit of a technical problem that might have required them to pay taxes, and so they developed a very technical solution that fixed it. The fact that the solution is a little shammy-looking would be a problem if everyone expected them to pay the taxes, but since people don't expect that, any old solution will do:
To people in the industry, heartbeats are just part of smart tax strategy. "It's removing a negative from the investment process," says Bruce Bond, a pioneering ETF executive who was among the first regular users of heartbeat trades. "There really isn't a need to have to pay the tax every year. The goal is the best investor experience you can have."
"There really isn't a need to have to pay the tax every year" is not something that you would say about your salary! But it's the sort of thing that an ETF guy can say about the ETF industry with a straight face, presumably because people-including many of the people who collect the taxes-really believe it.

Congratulations Lyft

Lyft Inc. priced its initial public offering last night at $72 per share, at the top of its $70-$72 revised marketing range and well above its $62-$68 original marketing range. I think you would call that a good result. It is only step one of the good result; step two is that Lyft will open for trading today, and a lot of other big private tech companies will be watching closely to see if it trades up. If Lyft's IPO prices above the top of the range and then trades well, then that is a very good sign for the next tech unicorn to go public. If Lyft crashes, then the party might be over as soon as it begins. There is something a bit irrational about this-if the IPO of the second-place ride-sharing company turns out to be overpriced (or underpriced), why should that mean that investors will stay away from (or flock to) the IPO of, like, an enterprise software company?-but I do not make the rules, and people really will be paying attention to Lyft's trading.

Bloomberg's Eric Newcomer interviewed Lyft's founders yesterday:
First, I wondered, would the company be profitable within five years?

The answer set the tone for the interview. Zimmer, Lyft's president, ducked: "We cannot talk about the future, but what we can tell you is that we have set ourselves up to deliver long-term shareholder value."
Here is a story about how Lyft has survived (so far) its bruising competition with Uber Technologies Inc. Here is an article about how the current crop of big tech unicorns are unusually unprofitable, compared to past big tech IPOs. And here is a fascinating metric from my Bloomberg Opinion colleague Shira Ovide, who compares big tech IPOs based on the ratio of (1) valuation at IPO to (2) actual capital raised in pre-IPO funding rounds:
Lyft has raised about $5 billion to fund its multiple business shifts and expansion. The IPO valuation looks to be about five times the amount investors had committed. Uber will go public at maybe five or six times its $20 billion in capital raised. Compare that with Facebook Inc., whose first-day public valuation was about 50 times the $2 billion in investments before its IPO. Google's IPO valuation was more than 600 times the money raised in private funding rounds.
I think this is the single most evocative metric for comparing the new economy of private tech companies to the old one. For one thing it's a good "private markets are the new public markets" statistic: Tech companies can stay private longer because they can more easily raise private money, whereas previous generations had to go public to raise real money. But it is also a good "venture capitalists subsidizing consumers" statistic, one that tells you what kinds of tech companies are big these days. Facebook and Google make easy-to-scale internetty things with low marginal cost; Lyft and Uber sell car rides at a loss. You need to raise a lot more money to do that!

Warren, what?

An occasional problem in modern stock markets is that, if you are buying a lot of stock in a public company, and you want to stay below 5 or 10 or whatever percent of the shares outstanding (to avoid triggering Securities and Exchange Commission filing requirements or a poison pill or exposure limits or whatever), you might not know what your actual limit is, because you might not know how many shares are actually outstanding. If the company has been buying back stock, that will reduce the number of shares outstanding (and increase your own percentage ownership), but you won't necessarily know how many shares they bought back in real time. (It's enough to make one want a blockchain, etc.) So this sort of mistake is entirely understandable:
Warren Buffett said Berkshire Hathaway Inc.'s investment in Delta Air Lines Inc. rose above his comfort threshold by mistake.

The stake climbed when Delta bought back its own stock and Berkshire increased its holdings, according to regulatory filings this month.
Then it gets a little weird though: "What I didn't realize was that that purchase had taken us over 10 percent," Buffett said Thursday in an interview with CNBC. "I was already in territory I didn't plan to get, so I just decided to buy a whole lot more stock."

"I mean, once I'd lost my virginity essentially, I thought, why stop at one, you know," he went on, because of course he did. The Oracle of Omaha! The Mayor of Graham-and-Doddsville! The greatest fundamental investor, and greatest explainer of fundamental investing, in history! And he goes on television to be like "oh yeah my investment rationale here was that I accidentally bought too much stock and I figured the way to fix that was to buy more stock."

Things happen

Wells Fargo CEO Abruptly Steps Down, Succumbing to Scandals. Wells Fargo's CEO Search: Here Are Likely Outside Candidates. JPMorgan's Role in Nigerian Oil Deal Has Come Back to Haunt It. Morgan Stanley President Kelleher to Retire at End of June. Hedging Rises on New Benchmark RateAmid Concerns of Quarter-End Swings. Goldman Sachs is exploring plans to create a Netflix for data, and it marks a new frontier for Wall Street. Ghost Villages Are for Sale in Spain. "For this week, what was important is a clean silhouette and classic pieces." "'Baby shark, doo doo doo doo doo doo,' reverberated through the ballpark."
  1. There is more to it than that but this is good enough for our purposes.
  2. The more useful way to think about it might be "ETFs do not sell *appreciated* stocks," but I will skip over that in the text. If you have to sell stocks that have gone down, generating a tax *loss*, this is all simpler.
  3. Meanwhile regular investors who want their money back really do just sell their ETF shares on the exchange; if there are more buyers than sellers, a market maker ends up owning a lot of shares and then giving them to an authorized participant to redeem.
  4. As the Bloomberg Businessweek article puts it: "In 1969, Congress decreed that mutual funds can hand over appreciated stocks to withdrawing investors without triggering a tax bill. Lawmakers never explained why they blessed the industry with this unique break, but back then, it didn't seem like much of a giveaway. Mutual funds rarely took advantage of it, because their investors preferred cash payouts.Fast forward to 1993, when American Stock Exchange executives devising the first U.S.exchange-traded fund realized they could put that old tax rule to a new use. Say an ETF needs to get rid of a stock that went up. Selling it would trigger a taxable gain that would ultimately be paid by fund investors. But handing the stock off to a broker who's making a withdrawal achieves the same goal, tax-free. Every time a broker redeems a stake in a fund, it's another chance to avoid taxes. That loophole gives ETFs a tax advantage over mutual funds. It went from a footnote in the tax code to the cornerstone of a new industry."
  5. I mean this is part of the explanation. Another way it works is that if you have stocks that have a tax loss, you sell those for cash. Another way it can work is that if you have realized tax losses-from selling losing stocks for cash-then you can also sell some winners for cash and shelter your gains with your losses.
  6. Why a few days? From the article: "Bankers prefer to tie up their capital for as short a time as possible. But ETF managers worry that too brief a holding period would cause the IRS to deem the transaction a sham. Years ago, some banks were completing heartbeats in less than a day, a turnaround time that alarmed ETF tax lawyers. Most funds now insist bankers keep their money in for at least a single day. 'The industry standard is 48 hours,' says James Brown, who counsels fund companies on taxes at Ropes & Gray in New York."
  7. "ETFs are typically more tax efficient in this regard than mutual funds because ETF shares are frequently redeemed in-kind by the Authorized Participants," says the Securities and Exchange Commission; the basic structure is not a secret from the authorities.