Tontines and Marshmallows

Also corrections, Somali shillings, bank bonuses, regulatory SWAT teams and unicorns.


Here is a fun story about the return of the tontine, the 17th-century financial innovation in which a bunch of people put money in a pot, agreed that whoever died last would get the whole pot, and then set about surreptitiously trying to murder each other and claim the prize. Or that is my general impression of tontines, and I am not alone: "Works by Agatha Christie, Robert Louis Stevenson and P. G. Wodehouse all featured tontine members plotting to kill one another in hope of a big payoff." (There's also a "Simpsons" episode.)

Tontines have mostly disappeared, but there is "a backward-looking cabal of financial specialists who suggest that this antiquated investment model may address modern challenges." Needless to say, one of them "envisions tontines layered with blockchain and smart contract technology." ("Mr. Brownstein said the technology could help track whether tontine members were actually alive or dead.") The key challenge that they're addressing is longevity risk, the risk that people might outlive their retirement savings. The tontine would address this challenge by creating incentives for people to kill each other quickly. No, no, I'm kidding! I don't think that's the idea?

The last year or so has been really hard on the notion that history is a linear progression from darkness to enlightenment, but it's been really good for theories of history as an endlessly recurring series of cycles. You see a bit of that in finance too: So much of "financial technology" -- bitcoin and blockchain and peer-to-peer lending -- is really about abandoning the modern technology of finance and returning to a simpler and more primitive time. (But with computers.) What if we got rid of the maturity-transformation function of banking and just loaned money to each other? What if we got rid of central banking and started mining our currency again? What if we got rid of securities regulation and just built a lot of pyramid schemes on a blockchain? None of this is driven by pure academic curiosity, or perverse destructiveness: It comes from a distrust of those modern institutions, a desire to return to olden times because the modern technology has somehow gone wrong.

The tontine revival seems to fit with that spirit. We had discovered perfectly good risk-sharing mechanisms to deal with longevity risk: Social Security, defined-benefit pensions, annuities. But trust in that system is breaking down, so we're going back to the past. Maybe 21st-century capitalism will eradicate 20th-century risk-sharing mechanisms like pensions and throw us back on 17th-century ones like tontines.

Marshmallow spoofing!

Christopher Mims tried to buy some marshmallows on Amazon, and fell into a world of bots. Also a world of surprisingly expensive marshmallows:

The vendor of the marshmallows I wanted told me his high price was an attempt to bait competitors into raising their own asking prices for the item. This works because sellers of commodity items on Amazon are constantly monitoring and updating their prices, sometimes hundreds of thousands of times a day across thousands of items, says Mr. Kaziukėnas. Most use “rules-based” pricing systems, which simply seek to match competitors’ prices or beat them by some small fraction. If those systems get into bidding wars, items offered by only a few sellers can suffer sudden price collapses—“flash crashes.”

I still don't understand how the marshmallow trade works? What good does it do you to raise your prices and bait competitors into raising their prices to just beat you? Is it a latency arbitrage -- you pull their prices up, then drop yours and get the sale before they can react? (Why would that work?) Or is it a sort of antitrust-conspiracy-by-bot, where you all raise your prices and get outsized profits because none of your bots wants to be the first to defect from the cartel?

There is something weird about a lot of discussions of high-frequency trading, on Amazon or on stock markets. The standard story about markets, generally, is that they are effective mechanisms for aggregating individually rational (that is: greedy) behavior into socially beneficial outcomes. The standard story about computers, generally, is that they lack many of our more appealing human qualities but are really good at cold calculating logic. You'd think that combining the two -- using computers to emphasize our most coldly rational and greedy qualities, and then using markets to aggregate those computers' individually hyper-rational behavior -- would work really well. I guess it does; that's why people keep doing it.

But so much of what you read about computerized markets suggests that all this aggregating of rationality leads to irrationality, that individually rational decisions by a bunch of computers tend, when aggregated by the market, to lead to collectively absurd results. (There are also stories about how individually irrational decisions by computers -- coding errors, fat fingers, etc. -- lead to absurd results, but those are less interesting.) Somehow a bunch of hyper-rational robots competing to offer the best prices on a transparent electronic market end up offering ... worse prices? Is that a failure of the market, or of the computers, or of the combination of the two? 

Would it be better for stock prices if stock prices went down?

I love this stuff:

What’s due now, some investors say, is a correction: a 10% pullback from the indexes’ March 1 highs. They contend such a retreat would tamp down speculation, deflate pockets of froth in popular investments and provide buying opportunities for those still on the sidelines.


“It’s like dental work,” said Michael Farr, president of the money management firm Farr, Miller & Washington. “You dread it. You don’t want to get it. But you’re glad when it’s over and you feel better.”

The subtext of most quotes in newspaper articles about stock prices is "I reject the efficient markets hypothesis." If you say that stocks are overvalued or undervalued or down on profit-taking or up on short-covering, you are explicitly or implicitly substituting your view of value for the market's. Which is fine! Lots of people reject the efficient markets hypothesis, and everyone loves a narrative.

But this goes one step further: It's not just that the market should be valuing companies more cheaply, but also that the market itself would feel better if it did. This is not just disagreeing with the market, it's condescending to the market; not arguing with the market, but telling the market that it's over-tired and cranky and would feel much better if it took a little nap.

The Somali shilling.

Here is J.P. Konig on the Somali shilling, "a paper currency without a central bank" since the Central Bank of Somalia stopped printing it in 1991:

Old legitimate 1000 shilling notes and newer counterfeit 1000 notes are worth about 4 U.S. cents each. Both types of shillings are fungible—or, put differently, they are accepted interchangeably in trade, despite the fact that it is easy to tell fakes apart from genuine notes. This is an odd thing for non-Somalis to get our heads around since for most of us, an obvious counterfeit is pretty much worthless. The exchange rate between dollars and Somali shillings is a floating one that is determined by the cost of printing new fake 1000 notes. For instance, if a would-be counterfeiter can find a currency printer, say in Switzerland, that will produce a decent knock off and ship it to Somalia for 2.5 U.S. cents each (which includes the cost of paper and ink), then notes will flood into Somalia until their purchasing power falls from 4 to 2.5 U.S. cents ... at which point counterfeiting is no longer profitable and the price level stabilizes.

Isn't that sort of how bitcoin works? But the central bank is working with the International Monetary Fund to re-introduce official currency:

Rather than repudiating counterfeits, the normal route taken by central bankers, the CBS will buy them up and cancel them. It will have to offer a decent price too, say like 5 or 6 U.S. cents for each 1000 note. If it makes a stink bid, say 3 U.S. cents, Somalis may simply ignore the appeal to bring in their old currency and keep using the old stuff. Because the buyback decision validates the work of counterfeiters, it just seems wrong. However, keep in mind that for the last twenty-five years it has been counterfeiters who have been willing to take on the risk of providing Somalis with a very real service, the provision of a working paper medium of exchange.

Well, but: What does it mean to be a "counterfeiter"? Classically, counterfeiters print notes that compete with notes printed by a central bank, hoping to capture some of the central bank's seignorage revenue. The Somali counterfeiters don't have a central bank to compete with, don't try very hard to imitate the "real" notes, and don't make much seignorage revenue either -- just enough to pay for the economic value they provide. 

Elsewhere: "With as many as 45 million British one-pound coins suspected fake, the Royal Mint is now making 1.5 billion new ones, which will enter circulation Tuesday." I suppose the goal is to make them cost more than a pound to counterfeit.


This is not the best look:

Some leading investors in Standard Chartered and Barclays object to plans to cut the profit triggers for long-term incentive plans to levels below the banks’ own stated targets.

StanChart has cut the lower threshold for its return on equity — a key benchmark of profitability — in its LTIP to 5 per cent, half the level of the bank’s long-term profit target. Some investors complain that chief executive Bill Winters will be rewarded for delivering an RoE below the bank’s cost of capital.

"Do we pay them a bonus for just coming to work," asks one investor, and of course the answer is yes. It's a conundrum. If you own a bad business, you still need to hire someone to manage it, and you need to give him the right incentives. If he achieves mediocrity, that is a big win for you; if you pay him only for excellence, he will have no incentive to shoot for mediocrity, and you'll end up with terribleness. Paying him a big bonus for achieving mediocrity is the right decision; it is embarrassing, but that is the sort of tradeoff that you end up making when you own a bad business. Amazingly Standard Chartered's 2016 incentive plan featured a bonus threshold of a 0.4 percent return on equity, and the managers missed it.


Here is an annoying article about how President Donald Trump's son-in-law Jared Kushner will be put in charge of a "White House Office of American Innovation," which will be "a SWAT team of strategic consultants," "an incubator of sleek transformation," and "an aggressive, nonideological ideas factory capable of attracting top talent from both inside and outside of government." The article features Marc Benioff saying that Kushner "does remind me of a lot of the young, scrappy entrepreneurs that I invest in in their 30s," presumably because he is in his 30s. And there is this, from Steve Schwarzman:

“There is a need to figure out what policies are adding friction to the system without accompanying it with significant benefits,” said Stephen A. Schwarzman, chief executive of the investment firm Blackstone Group. “It’s easy for the private sector to at least see where the friction is, and to do that very quickly and succinctly.”

That's a nice veiled description of Chesterton's fence: If you are trying to figure out what regulations impose costs without also providing any benefits, and you hire people from regulated industries who know about the costs ("at least see where the friction is") but who don't see the benefits, then they will tell you to get rid of the regulations. Ideally you would also talk to people who know where the benefits are. It's also a nice illustration of my theory that there are two kinds of regulation: Custom regulations, which are created to serve some specific goal; and bulk regulations, which add "friction to the system without accompanying it with significant benefits," are the subject of all discussions of "regulation," and do not exist. "It's easy for the private sector" to tell which regulations are pointless red tape, says Schwarzman, but you'll notice he doesn't name any.

Also Kushner is going to solve opioid abuse. And: "Trump to Issue Far-Reaching Reversal of Obama Climate Push." And: "BlackRock and Vanguard call for delay to fiduciary rule." And it's very easy for Carl Icahn, a "special adviser to the president" on regulation, to see which regulations cost his own businesses money, so he's going to get rid of them

People are worried about unicorns.

Here is Kadhim Shubber on liquidation preferences in later-stage venture capital:

When times are good, this sort of structuring gives investors downside protection and incentivises a founder to build as big a business as possible so they can sell at a price that dwarfs the preferred share stack, thus maximising their own payout.

But when a big exit looks less likely, you may end up with a founder who wonders why they are still losing sleep and weekends when they’re unlikely to get a look at the value they’re creating. They may have cashed out a little in earlier rounds and could have a sizeable financial cushion to land on if they simply walk.

This seems like the sort of problem that could be solved by renegotiating the liquidation preferences: You go to the late-series investors and say "look, I know I promised you the first $200 million of any sale price, but at current prices that doesn't really give me much incentive to work, so can we re-strike it to $100 million?" (Public companies do something very similar, repricing underwater stock options to keep their executives incentivized.) Obviously the investors won't be happy, but the problem here is the misaligned incentives, and fixing it can be good for both the investors and the founders. If you pay only for excellence, the founders will have no incentive to shoot for mediocrity, and at this stage mediocrity is what you want. I guess that's harder to accept at a tech startup than it is at a big bank.

Elsewhere, "Uber Group’s Visit to Seoul Escort Bar Sparked HR Complaint." And: "Uber to Suspend Autonomous Tests After Arizona Accident." 

Things happen.

Lynn Tilton: Diva of Distressed or Diva of Destruction? Inside Alabama’s Auto Jobs Boom: Cheap Wages, Little Training, Crushed Limbs. A German Bank Accidentally Transferred $5.4 Billion to Four Other Banks. North Korea’s Rising Ambition Seen in Bid to Breach Global Banks. Reviving Deutsche Bank in US is priority for CEO John Cryan. Poor Kweku Adoboli is fighting his deportation. OPEC Warns Members to Comply With Oil-Production Cuts. An Unexplained 85% Fall Is Nothing New for Hong Kong Stocks. A New Twist in iHeartMedia’s Debt Struggles. Scots shell companies used to launder £4 billion out of Russia. "We show that, over the life cycle, households progressively shift from growth to value as they become older and their balance sheets improve." "No Pay" Bylaws May Threaten Shareholder Lawsuits. "In the past week I have asked everyone I have met if they have read their employers’ codes of conduct. The majority either said no or looked sheepish and said they had skimmed it." ‘Fearless Girl’ will stay on Wall Street until 2018. Hashtag plane.

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Thanks! 

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

    To contact the author of this story:
    Matt Levine at

    To contact the editor responsible for this story:
    James Greiff at

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