In a recent Diff piece, we noted that Warren Buffett's takeover of Berkshire Hathaway started when he was annoyed that the company gave him some insider information to trade on and then changed its mind ($). Buffett asked the company if it was going to do a tender offer, the company said it would, and then when they did the offer, it was for 1/8th less than they'd previously said. So Buffett took over the company and fired the executive who had lied to him. This is a nice little corporate drama. But also, in modern terms, it's completely illegal: if Berkshire's CEO had tipped off a major shareholder about the timing and price of a pending tender offer, the company's general counsel would have rushed to file an 8-K, and Buffett's head of compliance would have insisted that he not trade the stock until the 8-K hit the wires.

But this isn't a story that anyone is particularly embarrassed about. The typical view, which is correct, is that Buffett was operating as expected under one set of rules, and when the rules changed, he adapted. He might have actually changed his approach a bit before that: there are anecdotes in The Snowball about other cases where executives revealed private information to him, which he traded on. He mentions it in at least one of the partnership letters, explaining that he tracked the progress of a company's liquidation by periodically calling them to ask when the next dividend would hit, and roughly what it would be. But these were small-scale opportunities. If you're taking a small position in a microcap, you might be the only investor management's talked to in years, so whatever they provide is probably alpha. But if you're carefully accumulating 13G-worthy positions in larger companies, knowing about some near-term catalyst just doesn't have much of an impact. What mattered when Buffett bought Apple was that he'd correctly judged how durable the business was, not his view on the next quarter.

There's been a general evolution towards higher standards for disclosure and behavior in US markets over time. On the disclosure side, there's a balance between keeping investors informed and forcing companies to give too much information to competitors: the more strict the disclosure regime, the more it operates as a tax on companies that choose to go public, and ensures that the most interesting and ambitious companies will stay private, concentrating gains with larger investors.¹

On the behavioral side, the picture is messier. Some activities are anecdotally less common than they used to be. Consider "banging the close," i.e. buying more of a position in the last few minutes of trading on the last day of the quarter or year, in order to make year-end numbers look slightly better. This was anecdotally common for a long time, and was basically a small wealth transfer from a fund's limited partners to its management company and that company's brokers. Turney Duff writes about doing it in The Buy Side.² It's hard to prove in any one instance—maybe someone spent the week after Christmas re-underwriting some of their biggest holdings, and happened to convince themselves of the bull case on all of them at 3:58pm on the last trading day of the year. But make it a pattern, and it's undeniable. If these trades were tolerated, they'd happen in a more adversarial market: since the net ownership of every stock is 100% long, there would be more upward pressure on the last day of the year than downward pressure, but traders would capitalize on this in advance.

Insider information actually fits into a similar category: different jurisdictions have different standards, both for what information can be traded on and by whom. The US, for example, frames it as a tipper-tippee relationship: whoever has that information and shares it is basically stealing it and selling it to the trader, who converts this stolen information into alpha. But this isn't the only way to set things up: Hong Kong, for example, doesn't care how you got the information as long as you traded on it when other people didn't have it. In the US, it's theoretically legal to make a trade based on a conversation between strangers about a pending merger that you overheard in an elevator; in Hong Kong, it's not. Both of these models work, and reduce bid/ask spreads by making it less likely that someone's making a last-minute trade based on misappropriated knowledge, but they set up different incentives. Sometimes, the rules are inconsistent across jurisdictions in a way that leads to unpleasant surprises on both sides: in The New Market Wizards, Jeff Yass tells a story about losing a lot of money very fast by selling a large number call options on a thinly-traded stock that got acquired ten minutes after Yass made the trade. As it turned out, a director at the acquiring company was buying the options: the acquiring business was from a different country where such trading was the norm, and had no idea that he was doing anything illegal. (It might have occurred to him that a country where you can make a trade that big, that fast, with an unknown counterparty, might be a country where that particular trading thesis was not a valid one.)

It used to be the norm that executives have flexibility to trade their stock based on their own information. In fact, there were a few cases in the late 1920s where they'd sponsor pools to manipulate their own company's shares higher. This was seen as an obvious perk, and was part of how executives got paid. And, in an environment where record-keeping is inconsistent and trade surveillance is difficult, it's hard to ban this kind of transaction. But what it amounted to was that directors got paid a variable amount, based less on how much value they added to the board and more on how much value they could extract from shareholders who didn't have boardroom access. What this selects for is a population of directors who are good at trading but not necessarily good at business, and who actively seek out the most volatile companies because they'll have the most exploitable news flow. It's a lot easier to give them fees and stock options instead.

But it's important to recognize that all of these regimes set up equilibria, and sometimes surprising ones. If you wanted to shrink the financial sector and reduce the amount of human capital invested in spotting and exploiting stock market anomalies, an effective way to do this would be to legalize insider trading. A narrow spread in a stock is, implicitly, a bet that traders on both sides are unlikely to have material information that will move the price quickly; change the odds of that, and the optimal bid/ask spread gets a lot wider. But robust capital markets are a net good thing, both because they move capital and because they transmit information so fast. As it turns out, one way to improve their information transmission abilities is to block some informed people from letting the market reflect what they know.

Read More in The Diff

In The Diff, we’ve covered regulatory regimes and their tradeoffs in many contexts:

• We’ve made the case for marking more things to market.

• We’ve looked at the economics of noncompetes, and which workers are harmed and which actually benefit from them ($).

• Equity compensation is partly a story of regulatory arbitrage that evolved into a story about incentives.

While this piece closes by noting that we could shrink the financial sector by loosening some rules, that might not be a good thing after all ($).

Join the discussion!

1  In a way, this is not such a big difference, because the LPs for a typical venture or growth fund will include plenty of pensions, endowments, etc., so in the end the capital gains accrue to some sympathetic figures. But those capital gains, unlike public market gains from an index fund or other cheap vehicle, come after carry: the larger the delta between public and private company disclosure requirements, the more public company returns are converted into private equity fees.

2  In Duff's story, he would tell his broker which stocks he planned to buy a few hours in advance, but tell them not to make the trade until just before the close. He noticed that some of these stocks would suspiciously drift up before his bid—his broker was leaking the plan to someone else! So, he laid a trap: he did exactly the same routine the next year, but, through a different broker, shorted the stocks in question. And then, just before his order was set to execute, he called up the original broker and canceled it. His broker was factually correct that it's easier to rip someone off if they're doing something unethical, since it's not as if Duff was going to complain to the SEC that his broker was interfering with his market manipulation. But there's also a pragmatic case that if you spot someone doing something a bit dubious, and decide to take advantage of them, you're competing with someone who has already displayed skill at executing minor rip-offs, and who presumably doesn't think they're the only sharp actor in the market.