Byrne's Eye View

The other day, I read Tim Mackintosh-Smith's Yemen, which is a delightful book -- sort of what this would sound like if it had been written by a sociology geek instead of an engineering geek. This struck me as especially odd, though:

South Aravua exported an estimated 3000 tones if incense and 600 tons of myrrh annually. Given that the people of Rome alone spent 85 tons of coined silver a year on incense, that myrrh was vastly more expensive, and that the spices and other luxury goods which passed in transit through South Arabia fetched similarly high prics, the income for the Sabaeans and their neighbours would have compared favorably with the present-day revenues of an oil-exporting state.

Resource disparities have been leading to massive, unpredictable wealth transfers for thousands of years, but unless you can engineer a massive culture transfer, too, the same kind people will get rich regardless of which resources they can exploit. It's fascinating that US trade policy and immigration policy focus on keeping resources flowing in and smart people out -- if you ever want to compare closed society to open society, keep in mind that the vast majority of Korea's natural resources are in North Korea (which is also the first country since the Mongols to fund their trade deficit by exporting threats of violence).

This Fortune profile of Robin Hanson is a pretty good overview of the early history of prediction markets, but it makes a few obnoxious errors:

There is no telling whether [the Policy Analysis Market, the prediction market that focused on, among other things, potential terrorist activity] would have worked. Experimental economics veterans Plott and Forsythe are skeptical. "When you have a very small number of people with very closely held information, I have some doubts about that kind of a market," Forsythe says. Another potential problem is something economists call "moral hazard." It wasn't that, as some PAM critics claimed, the market would provide terrorists with a monetary incentive to carry out attacks: In its initial phase PAM was going to be restricted to a few hundred participants drawn mostly from the intelligence and foreign policy establishment, and the maximum investment allowed would have been $100. The real moral hazard with PAM was the risk that an important intelligence analyst might manipulate the market by changing his own forecasts.

Any time you have a measurable, quantifiable risk, you can let people hedge it through prediciton markets. If everyone is reading the same five intelligence sources, and each source is also playing the markets, all you need to do is let people bet on the reliability of each source, too. At a pretty low cost, you can simulate any level of reliability (i.e. if a 60% reliable source says that there's a 60% chance of another major terrorist attack in the next year, you'd expect the next-year-attack contract to be worth .60 * .60, or $.36 -- but if you bet in favor of the attack and against the reliability of the source, you've used securities markets to synthesize a 100%-reliable source).

It's another sign of superficial understanding when a writer can discuss a subject, bring up a criticism, and not realize that what he's discussing provides a framework for solving the problem he's noticed.

Edge.org asked a few leading intellectuals for pithy summaries of what they believe but cannot prove. I liked Nassim Nicholas Taleb's:

We are good at fitting explanations to the past, all the while living in the illusion of understanding the dynamics of history.

There's a lot of effort devoted to cargo cult investing -- replicating the investing process of the past in the hope that it duplicates past results. But what's more interesting from the perspective of Taleb's theories is that success in finance is harder to spot because once you can describe it, you can mimick it, and once you can mimick it, it doesn't work. So if you turn the study of investment models into a science, your only test subjects will be the mediocre.

Monty's Bluff on reputation:

Can you imagine if there was a person whose full time job was to criticize every action you took! [Did you see what John was wearing today? Does he actually think that is a nice shirt! Look, now he is eating a donut! Doesn’t he realize what a strain he is placing on the nation’s medical system?]

This is probably the best explanation for why celebrities -- professional athletes, actors, and media-exposed CEO's -- make so much more than the rest of us: once you're synonymous with a product or a company, you're giving up all of your privacy and free time to maintain your product's value.

The Wall Street Journal reports that Michael Oxley has been named Vice Chairman of the Nasdaq ($), which is, if nothing else, pleasantly symmetrical.

Dealbreaker has an exclusive scoop -- unless they're kidding -- about how much Sarbanes-Oxley is costing Warren Buffett.

According to Buffett, Berkshire-Hathaway spent $24 million on auditing this year, a figure he says would have been closer to $10 million without Sarbanes-Oxley. This is just one, anecdotal data point but still. That's an enormous cost multiplyer. If SarbOx is costing anything like this with many other companies, we're looking at a serious wealth transfer because of a regulation whose benefits are hard to quantify. Does anyone really think SarbOx is making investors 2.4 times safer?

Useful for train-wreck syndrome purposes only: a quick rundown of who lost how much in the subprime mortgage collapse.

The collapse itself isn't too worrisome, though: from what I've seen, the market is pretending that defaults among overleveraged speculators imply that higher-grade debts are next. The easy money (defined as "money that will run out shortly") has run out, but the real estate market as a whole is solid.

reddit has recently been all-atwitter over the FizzBuzz problem: most developers, when told to write an elementary program, simply couldn't. They weren't tested on their understanding of algorithms or their memorization of syntax -- they just needed to know how to express modular arithmetic in their language of choice. It looks like a great way to separate the real programmers from the cargo cultists, and it leads to another interesting question: what kind of simple FizzBuzz tests can we use in other disciplines?

I suggest that a FizzBuzz test for economics is a person's behavior on escalators. Anyone who has studied economics knows that rational self-interest dictates that we continue to take an action until its marginal cost reaches its marginal benefit. But only someone who really groks economics realizes that this makes standing still on an escalator as ridiculous as standing still anywhere else: if the effort of walking is worth getting you to your destination, it's worth it whether your base speed is 0 or 5 mph or negative. I suspect that the first-year undergrad econ students who continue to walk even when on an escalator or an airport conveyor belt are much more likely to finish their coursework and go to grad school, because they've so subsumed their economics knowledge that they can instinctively apply it to situations involving no graphs whatsoever.

Of course, that's only one of the ten principles of economics, so interested parties are encouraged to suggest more economics FizzBuzzes.

The WSJ reports that Delta Financial will keep its three-letter symbol even after transferring to the Nasdaq ($). Confusing, and sinister: four-letter symbols are a powerful branding mechanism for the Nasdaq, and the notion that they'd forgo that consistent distinction to make just one more deal is distressing. Exchange listings are becoming a commodity -- which means it's only a matter of time before they're priced like commodities. Can't be good for the Nasdaq or the NYSE.

It's pretty well-known that most financial advisors underperform the market, and it's obvious that this doesn't dissuade people from hiring them anyway. There's an easy conclusion to draw here: people are idiots who don't know what they really want. There's also a harder conclusion: people are making a more nuanced decision than a simple choice to pay extra fees in order to earn less money.

I propose that there are three basic kinds of investment managers, roughly categorized into how often they're (market-beatingly) right, and that each one provides a distinct service:

• Those who are right 49% of the time are not in the business of managing money. They're therapists for a generation whose neuroses are mostly financial. As the print, radio, and televised financial media become more ubiquitous, they promote an unrealistic image of the market to a typical investor: exaggerating both the prospect of gains and the danger of losses -- treating the market as a game that everyone ought to play and that we're all required to be good at. This isn't surprising, since that's the most entertaining way to portray the market, and entertainment is what the media need to offer. Unfortunately, it convinces many people that they ought to be in the market and that they ought to be ashamed when they lose money, which isn't a pleasant state of affairs. For these people, it makes sense to let someone else run the money and take the blame: if an unskilled investor runs his own portfolio, losing money exacts a financial and psychological toll -- if he pays a manager, the financial toll is higher but the psychological cost is minimized. I suspect that this is a fair deal for a large fraction of the investing public, because that's the only way to explain why so many of them accept it.

• Advisors who are right 50% of the time are in the business of aggressively not screwing up: to the extent that they can match the indices before fees and avoid any egregious innovations, they'll keep making money. This was obvious for a long time, but Vanguard was the first company to admit it, and, as a consequence, the first company to make serious money at it -- in much the same way that Apple was the first company to really admit that computers were intrinsically fun, creative, anti-establishment tools, rather than the Capital Expenditure of Choice for Avatars of the Establishment.

• Advisors who are right 51% of the time are easy to envy, because they get to be honest about their jobs: they legitimately try to think differently, to outguess their competitors, to efficiently structure a portfolio so taxes and transaction costs aren't a serious burden, and to allocate capital to where it can generate the highest return. All very exciting to the kind of person prone to getting excited about such things, and thus brutally competitive and rather difficult to do adequately. Picking a '51%' manager requires examining a track record, cross-examining shareholder letters and past portfolios, and estimating the relative roles of luck, skill, and circumstance. In short, picking a good manager well is about as easy as picking the stocks on your own -- '51%' managers use the same rhetoric as most '49%' and '50%' managers, and there isn't an effective algorithm for deciding who has genuine ability and who is about to painfully revert to the mean.

You can already tell where this is going: for someone who values their time and their sanity, picking an objectively worse advisor is probably a better deal, and the people who can pick an intelligent fund manager are part of the same demographic that can pick their own stocks. The equation is more than just returns minus fees adjusted for volatility: it's returns, sanity, and the pursuit of (a unique and idiosyncratic form of) happiness.

Michael Oxley, as in Sarbanes-Oxley, on Sarbanes-Oxley:

"Frankly, I would have written it differently, and he would have written it differently. But it was not normal times."

Amazing! What would prompt an elected official to be so humbly insightful about the effects of overreacting based on emotional responses to limited data?

Presiding over a recent dinner in Paris for more than 200 accountants, Oxley — the former Republican congressman from Ohio...

Oh. Right.

This is odd: Bats Trading, a small exchange that is now third after the NYSE and Nasdaq, managed to trade last week without any of the glitches suffered by the larger exchanges. Perhaps the network effect is overestimated for exchanges: if each exchange trades the others' stocks at cost, it's pretty hard for size to be an advantage.

And speaking of recessions and bear markets, the New Yorker's James Surowieki parses the panic last week and believes that 1) it was irrational, 2) but it was also rational, and 3) either way, that's pretty much okay.

I can never get interested enough to make a big deal about one-day declines: I'm not leveraged now, and I can't imagine being leveraged enough for a move like that to wipe me out, so it would be a buying opportunity if anything -- but it's hard to treat a single-digit percentage move as a serious opportunity, either.

Abnormal Returns notes that during a bear market, correlations between markets tend towards 1, which makes theoretical sense: when the economy is growing, it's growing in lots of different ways, but when it's shrinking it's mostly shrinking due to a reduced money supply, which affects all firms similarly.

The only companies that aren't harmed by a bear market are companies that have essentially shorted liquidity by buying only short-term obligations and lending out long-term -- but in general, a serious bear market will only happen when such companies have capitulated.

Here's an idea: instead of rewarding research by granting temporary monopolies (patents), why not reward research by paying a bounty on it? Economist Joseph Stiglitz likes the plan, and he's a smart guy so I'd like to agree with him.

Unfortunately, there's a pretty serious objection: a patent amounts to a prize for research, awarded by taxing the beneficiaries of the research (there's no difference between a completely free market with a subsidy to discoverers, funded by an excise tax on the new discovery). The bounty system works like that, but distributes the tax burden to everyone. This doesn't make sense -- there's a whole class of scientific discoveries that I don't care to fund and don't even want to happen, and there's another class that I'd gladly pay for the benefits of but would hate to force someone else to subsidize. Stiglitz's proposal is to create a new bureaucracy specifically to muddle the question of costs and benefits. It's an expensive and risky way to break the market.

The stock exchange boom-or-bubble proceeds apace: The Bombay Stock Exchange is valued at about $860 million. But what I really like is this:

and the exchange in Calcutta said it planned to sell 51 percent of its shares to partners and strategic investors in increments of about 5 percent each.

The point of being a 'strategic' investor, as opposed to a passive investor, is that you get information and a chance to nudge the firm's growth in a helpful direction. Apparently they're going in ten different directions while creating ten exclusive channels for valuable non-public data.

Or they're saying whatever it takes to turn outsider optimism into insider cash.

A company called WebMoney, which from what I can gather is an Eastern European paypalesque enterprise, is issuing a currency that seriously competes with the Belorussian Ruble. Belorussia is not exactly a bellweather economy, but it's still mighty amusing to see private enterprise sidestepping the government to come up with their own entrepreneurial monetary policy.

As usual, good science fiction is a surprisingly good guide to what's coming next.

That's pretty brazen: Fortress Investment Group's net worth is now negative because of the massive cost of paying bonuses to the principals -- who cashed out as the company went public anyway.

It's legal, and it should be legal, but it's a good reason to avoid buying their stock or dealing with their funds: if this is their idea of fiduciary responsibility to their shareholders, just imagine how they treat their limited partners.

Felix Simon is looking for activist mutual funds. I don't think there are any, because mutual funds have been regulated out of that business: as far as I know, they can't borrow much, and they can't own more than 10% of a single company. Activism requires that the activist 1) buy and be able to hold indefinitely, and usually 2) be able to borrow enough money to buy a controlling stake. When you have limited leverage and your available capital is a function of how good your PR team is, that's not likely.

I could be wrong, though.

Roger Ehrenberg writes about Apple's cash hoard and what ought to happen to it, and concludes that Apple is in as bad a position as having $12 billion can put you in: the marginal return from doing Apple-y things is low, but if they do un-Apple-y things it'll contrain their ability to remain Apple-y. Fortunately, they can just return capital to shareholders.

Except to the extent that increasingly volatile business conditions mean that they can't.

This brings up an interesting question: as derivatives and prediction markets proliferate, why do we see more expected volatility in returns, rather than less? Shouldn't two-sided volatility (component costs, for example -- when they change, one party benefits at another party's expense) be something close to nonexistent when you can bet on them in advance? There are plenty of speculators with the resources to take the opposite side of otherwise counterparty-free trades, so I have to wonder just where all that effort and money going into derivatives has ended up.

What's the highest instantaneous real gain the stock market as a whole can produce?

I'll use one-day price change as a proxy for this: the highest one-day gain was 14.87% in 1931; the largest one-day loss was 22.5% in 1987. This itself is pretty surprising -- it shouldn't be possible for a reasonably efficient market to spontaneously decide that the future earnings of American industry are 15% higher or 23% lower overnight. Markets are certainly driven by more than just fundamental analysis (both moves were technical, at least in the sense that they were driven by idiosyncrasies in how trades are executed, rather than by economic factors), but technical problems are, by their nature, hard to fix -- NYSE and NASDAQ are two brutally competitive companies whose business models each involve being more reliable than the other, so if anyone out there is going to put in the effort to predict a market breakdown, they're probably also going to fix it in time.

But fundamental moves are still a rich source of interesting speculation. For the market to make a massive upward move, there would need to be a massive, instantaneous macro change -- and the only source of those is government.¹ But it's hard to imagine a government edict having a huge positive effect on prices: any change in income distribution (lowering corporate tax rates, etc.) will also lead to a corresponding change in spending or deficit, and since the people paying the difference are all consumers, employees, and shareholders, the changes would be redistributed pretty fast -- because different companies have different pricing power, it wouldn't be even from one stock to the next, but for the market as a whole, even a radical redistribution of wealth to publicly-traded companies wouldn't have a huge net effect (besides decreasing the losses due to tax collection and obedience to regulation).

It probably takes a lot of imagination to dream up real gains for the stock market because politicians have every incentive to make them happen -- if there's a policy with few to no ill effects, but which in general makes people slightly richer, Congress will clamber for a few more percentage points of annual gains to take credit for. They've pushed the limits of easy gains (free trade agreements, SEC regulations keeping investors informed, a generally pro-growth tax-and-regulation policy) and are bumping into the no-win tradeoffs of redistributing one sector's outsize gains to offset another sector's extra losses (via ethanol subsidies, gas taxes, and other such well-meaning mistakes).

What's interesting is that it takes little effort at all to imagine a policy that could suddenly devalue the market. Anything from a loopier tax policy to a more aggressive foreign policy to an inefficient attitude towards trading and capital gains could send the market down to an arbitrarily low level. Pick a loss -- 5%, 25%, even 95% -- and it takes minimal effort to dream up a series of missteps that could get us there. It ought to be worrisome that a Dow 13,000 tomorrow is impossible, but a Dow 1,300 tomorrow is one late-night vote away. A single senator or representative couldn't hope to generate a 1% rise in stock values, but could easily aspire to a 10% loss. In brief:

The healthiest political ideology for the market and the country is simple: a political platform without a single plank.

[1] The only other one would be a huge, paradigm-altering change in technology, but the whole point of paradigms is that they're beliefs held so strongly that people act on them -- and it's hard to imagine a large number of people spontaneously changing a belief that's central to their lives. Government edicts, on the other hand, change minds pretty fast because they can't be opted out of.

This fascinating article by Greg Robb at Marketwatch argues that Alan Greenspan should not say negative things about the economy because 1) everyone treats everything he says as infallible, even when they hear it second- or third-hand, and 2) someone could lie about what they heard Alan Greenspan say.

Interesting. You know, I heard Alan Greenspan say that Greg Robb should take into account the fact that people evaluate second- or third- or nth-hand claims by considering the credibility and motives of everyone involved, and by questioning the plausibility of the claim itself.

I forgot to mention this earlier, but Berkshire Hathaway's 2006 annual report is out (in .pdf format). Unfortunately, it's not nearly as information-dense as previous versions have been. The report definitely has enough data to keep current shareholders happy, but for the first time in a while it's not worth reading for outsiders.

Panic, Bubble: Market in which the main news driving the market's performance is news of the market's performance.

If this isn't coincidence, it's hilariously ironic: an insider trading scheme centered around Morgan Stanley and UBS has just been busted. Morgan Stanley has been falling for days -- before the story broke.

There's a precedent, as always: Ivan Boesky briefly considered shorting index futures before the official announcement that he was guilty of insider trading -- had he stayed short, he would have recouped a large portion of his fine.