Economics and similar, for the sleep-deprived
A subtle change has been made to the comments links, so they no longer pop up. Does this in any way help with the problem about comments not appearing on permalinked posts, readers?
Update: seemingly not
Update: Oh yeah!
Monday, October 31, 2011
On Liquidity
By way of an initial reply to Brad's Prologomenon:
The saver has his greed and has his fear
He wants it in an overnight account.
The borrower, though, wants the full amount
and wants to use it for a dozen years.
And so a broker magically appears
to offer stock, and bid at a small discount
and make a turn; but they must never miscount.
Liquidity providers cannot clear
a market which displays adverse selection.
Smart traders bearing lemons cause a drought.
Markets don't just exist; they must be made
So dealers churn their books in self protection
And gamblers wash the information out.
They also serve who only sit and trade
What am I on about? Well, basically the "No trade theorem" of Milgrom and Stokey.
Basically, if all you have are informed traders, then you don't have a market. It's similar to the market-for-lemons problem - the simple act of offering to sell a stock reveals that you have information that there's something wrong with it, and so nobody is going to trade with you. You have to have at least some traders who are motivated by liquidity considerations (ie, they have a gas bill to pay so they are selling stock, or they're just received an inheritance so they're buying) or who are just "noise traders", for the whole thing to work.
This is market microstructure - the study of *exactly* how trading progresses between informed parties in securities markets. Basically the conclusion is that market-makers (the professional intermediaries) lose money when they deal with informed counterparties, and make money when they deal with uninformed ones. And, since informed counterparties do not wear a badge identifying themselves as people to avoid, basically all the insurance industry apparatus of adverse selection can be theoretically brought to bear; as with the excess on an insurance policy, a large proportion of the cost of dealing in shares is basically the result of the need to overcome an information asymmetry.
I don't think I'd be exaggerating if I said that more than fifty per cent of the salaries and bonuses paid on a trading floor are basically to do with the business of making sure that you don't lose more trading with smart punters than you make dealing with mug punters. There are certain hedge funds that you always lose money when you take the other side of their trades - you deal with them a) because they rebate some of the profits they make out of you in the form of a commission payment, and b) because having their order flow means that you have some of their information, which means that you are now a smart counterparty with respect to the rest of the market, for a while.
And this is another reason why dealing in emerging market securities is so damnably expensive - the markets are smaller, and in general the only people dealing in them are people who have a reason to believe that they know what they're doing. Who are, of course, exactly the kind of people you don't want as your counterparty.
As far as I can tell, Brad's view is that there are natural, good liquidity traders (basically life cycle savers), and that this ought to be enough in the way of uninformed trading for anyone to be getting on with. I don't think that can be assumed at all though. Liquidity - the ability to have instant access to your savings even though they are invested in long-term capital projects - is a really valuable and popular thing, and people are prepared to pay a lot for it. And if it's a really valuable thing, then the market will find a way of paying for it (the original DeLong//Shleifer/Summers "noise traders" paper showed how without intending to, the market can find a way to pay positive feedback gamblers for their liquidity). I would not want to defend any particular level of mindless gambling as optimal, but some is better than none, and I don't think it can be flagged as a per se undesirable.
(also - measuring "good" trades as a proportion of the total turnover is a screwed metric. Most of the volume on any given day in nearly any given market is inter-dealer, and simply reflects the annoying tendency of end-users to turn up placing orders that are larger than the risk-bearing capacity of the firm that they placed them with. Inter-dealer volume of this sort is usually uninformed, except in those cases like the hedge fund I mentioned above when it is. But informed or otherwise, it's just necessary inventory management of the sort you get in any market and should be exed out of the market size for purposes of deciding how much trading is of what type).
this item posted by the management 10/31/2011 02:56:00 PM
Friday, October 21, 2011
The lost cure for cancer does not lie in a hedge fund somewhere
For some reason, a bit of a blast from the past is on my mind; Emanuel Derman's book "My Life As A Quant".
In summary, this book reads: "Once upon a time I was born, grew up and did a physics PhD. I enjoyed it, but it was pretty clear pretty quickly that I was no better than a mediocre physicist and had basically no chance of a prestigious academic job. So I went to Bell Labs, where I was a workmanlike professional scientist and computer programmer, but my career wasn't going anywhere there either. Then I went to Goldman Sachs, and everyone regarded me as a genius!"
Now it could be the case that Derman was a late bloomer and suddenly had an explosion of intellectual talent. It could be, but actually it wasn't; fair do's to the guy, he makes it abundantly clear in his autobiography that Fisher Black was the genius and that he himself spent an inordinate amount of time doing stuff on user interfaces that would be farmed out to a Visual Basic studio in Chennai today. But, it has to be admitted, that by the standards of investment banking, Derman was a genuine intellectual hero.
Basically where I'm going with this is that there is no great suction of the very best talents and most gifted of youths into securities trading, and no consequent intellectual peace dividend that could be reaped from regulating them. Yes even at those magical high frequency trading firms.
I've known quite a few guys who had physics PhDs over the years and a lot of them were pretty good. But in general, the ones who were good were also pretty realistic about whether they had the talent or aptitude for a life in physics. What the physics quants were always very good at was the systematic application of a mathematical toolkit to a problem, and very hard work on intrinsically not very interesting and not very cutting-edge problems. Lots of them also had the most amazing trainspotterish memories for detail; it's only a small exaggeration to say that a very large part of the design of algorithmic trading systems is the management of adjustments for bank holidays. They're the journeymen of physics, the builders of lab equipment, not the Einsteins. Most people in investment banking are either smoothy-chops sales types who might as well be there, or human wrecks like me that are pretty much incapable of anything else.
this item posted by the management 10/21/2011 07:40:00 AM
Thursday, October 20, 2011
Subtle shifts in the Party line
Apparently the Daily Mail is no longer as keen on foreign-born athletes competing under the British flag in the Olympics as it used to be.
this item posted by the management 10/20/2011 04:38:00 AM
Tuesday, October 18, 2011
Rarely is the question asked - is our children welding
Via Dean Baker, Thomas "Airmiles" Friedman reports on the travails of Caterpillar, Inc:
"We cannot find qualified hourly production people, and, for that matter, many technical, engineering service technicians, and even welders, and it is hurting our manufacturing base in the United States. The education system in the United States basically has failed them, and we have to retrain every person we hire."
Forced - forced! - to provide training for skilled staff! The American education system has failed - failed I say! - to provide an army of skilled welders and engineering service technicians, all lined up three deep and ready to work at whatever the CEO of Caterpillar decides is the going wage. What a total, total crock. The CEO really honestly believes that the idea of training his employees is bizarre behaviour akin to sexual perversion. No wonder the American teaching unions are asking for more money, if they're being asked to drag arc-welding kit from classroom to classroom, along with the textbooks.
this item posted by the management 10/18/2011 03:08:00 PM
Violence: The Great Moderation, with all that implies
I too haven't and (barring major advances in gerontology) won't read Stephen Pinker's book on violence. In general, I am allergic to Pinker's tone (I suspect that many of the people who, according to him, are always mocking him for not knowing about Milton, are actually just asking him to stop acting such a twat), and you can usually find the same views that he expresses written up much less annoyingly elsewhere.
In the case of the violence book, however, and rather hilariously, if I am correct in understanding that his main thesis is that modern societies have developed more effective methods of social control over the years, and that these advances in social conditioning have reduced the incidence of physical violence, then I know where the less-irritating writeup is - it's in "Discipline and Punish", by Michel Foucault. This is actually quite a common trait in New Rationalist writing, by the way - a lot of the time, while they will take pot shots and swipes at their ideological enemies, the real vitriol is reserved for their competitors (after all, enemies can make up, but competitors can't). Sigmund Freud said right back in the nineteenth century that a lot of human behaviour is caused by basic animal drives that have their root in Darwinian evolution, but my god if you read Dawkins or Dennett, you'll be aware that Freud is considered a bounder these days by people who believe the same thing.[1]
We are of course much less violent now, we have robotic drones to murder people from the sky rather than fighting them ... which raises another question in my mind. It's true, as I have written here before, that in many ways it was a fantastic achievement that the West managed to defeat Soviet communism without firing a shot [2], and that actually in many ways a lot of the credit for that deserves to go to those nemeses of mankind, economists.
But hell - although it is true that strategic deterrence worked, the way in which it worked was to spend forty years with both sides literally threatening something close to the complete destruction of life on earth! Which if it had happened, would have been considered by the survivors as a pretty violent thing to do.
Fair enough, as I say, strategic deterrence worked. But, sample size of one here? It might not have worked. It certainly did involve taking a risk of something very bad happening. And I think what we've all learned from the last five years is that when you take a risk that something bad might happen, there is a risk that something very bad actually will happen. I think I will, rather gratuitously, give Pinker a small dose of his own medicine, by suggesting that it is fairly obvious that a risk of a very horrible amount of violence, is clearly equivalent to a proportionately smaller amount of violence, and that it is disgusting, shameful, comical, etc etc etc, that someone who looks down his nose at you for not having heard of Maynard Smith apparently hasn't bothered to look up the relevant chapter of Schelling.
[1] It is true that Freud took this basic insight and went with it to a lot of rather ridiculous places and some fairly disgusting sexual politics, but evolutionary psychology is not exactly in the best position to comment.
[2] By which I mean "without having another massive war in Europe"; the objection about secondary theatres and Third World conflicts was well made (although not necessarily totally agreed with by me) last time round.
this item posted by the management 10/18/2011 05:05:00 AM
Friday, October 14, 2011
Wingnut welfare benefit tourist
I was discussing this on Twitter with Jamie Kenny last night - Liam Fox does indeed have some serious questions to answer to the British public, and the first one he needs to answer is “do you think we're all simple or something?". If a mate of mine happened to bump into me when I was on a business trip by coincidence, and we met up for dinner and he brought along someone who I didn't know but who was a contractor lobbying for a huge amount of work with my firm, my reaction would not be "cor, great call, let's do this another forty times, assuming that forty more purely coincidental meetings happen". It would be "mate, love to the wife and everything, but Christ, don't put me in that sort of embarrassing position please". If (per impossibile, because apparently my mates have rather more class than Liam Fox's) I had a mate who I suspected was pimping out his connection to me for commercial advantage then … well, I don't know what I'd do because this would never happen, but I like to think I wouldn't sleepwalk into destroying my reputation.
So what gives, if we presume that they're not lovers, and that the innuendo in the press is dancing round the accusation nobody dares to make yet - that these meetings actually were what they very much look like. In other words, Good Old British Defence Procurement, the process whereby troops in the field are used as hostages by the defence industry to extract a ransom from the taxpayers. I presumed that it was.
Jamie's theory is a little more interesting, and I'm getting a little more convinced by it, as my original reason for rejecting it - that it's totally insane to pay that much money for access to Liam Fox unless you're on the defence procurement gravy train - doesn't necessarily hold water. It matters, I think, that Fox is specifically an Atlanticist and proprietor of the Potemkin thinktank par excellence "The Atlantic Bridge".
Because, of course, as one of the earliest posts to this blog noted, while the cash to support Adam Werrity's lifestyle is a crazy amount of money to pay for access to a minor politician in Britain, in America they do things rather differently. Since Fox is an actual cabinet minister, I can see how an American billionaire (ie, someone potentially in the habit of dropping large five figures on Senate races) might regard him as compellingly good value.
In other words, Liam Fox may be the first guy to have successfully arbitraged the wildly different valuation of politicians in Britain and America. And the fact that he was so keen on the "Atlantic Bridge" suggests to me that this trade might have been on his mind for quite some time. In any case, now that the gate's been opened between the British pool of labour and the American pool of wingnut welfare, you can bet that people are going to work it, particularly people who work for a British thinktank that is, almost unaccountably, named after an American politician. Lower tier wingnut pundits in the USA like Jonah Goldberg or Michelle Malkin ought to be very concerned indeed that they are about to face a wave of the pundit equivalent of Polish plumbers, prepared to churn out a somewhat higher quality of boilerplate hackery for much lower wages. And with Murdoch-level ethical standards too. The world is, as Thomas Friedman said while earning approximately twice as much as Martin Wolf, flat.
this item posted by the management 10/14/2011 01:50:00 AM
Tuesday, October 11, 2011
Hmmm, not a vintage Krugman day
I find myself disgreeing with two posts, oddly:
First. It appears to be my role to be the Sisyphus of people trying to rewrite recent financial history to make the narrative more ethically straightforward. The benefits of EMU were not "equally shared between the core and periphery". Have we forgotten the Celtic Tiger so soon? What about "Old Europe"? There really was no massive benefit to France and Germany to remotely compare with the experience of Spain, Ireland and Greece (Italy, as always, falls half way between stools).
Germany, in particular, can't be expected to follow this sort of reasoning. A forty year old German born in the west has credible reason to believe that his entire working life up until 2008 had been spent paying for the cost of currency unions that mainly benefited other people. An forty year old German born in the East has had about five minutes' time in the sun before being told that the party's over and it's time to pay for the booze that someone else drank. Ireland still has higher per capita GNP and median income than France, by the way.
The case for Germany and France picking up the bill for keeping the single currency together, and alleviating the costs of Greece servicing its debt burden, is powerful. But it's never going to be made if it's to depend on the French and German electorate being somehow convinced that they profited out of EMU, at the expense of Greece and Ireland. Is it? Once more, as I said before in a more obscure venue, the case for solidarity and egalitarianism has to be based on solidarity and egalitarianism, not on some idea of theft.
The second one is more of a quibble; Prof K is basically right about all of this, but it's a real shame that he's citing Diamond & Dybvig (1983), rather than "Lombard Street", which is a great book.
this item posted by the management 10/11/2011 09:06:00 AM
Saturday, October 08, 2011
Reading between the lines ...
... often lets you know how well the journalist got on with his interview subject. Here's a beauty from this interview with Nicholas Taleb's old head trader, now running his own fund:
"Most of the time, you look like an idiot," says Spitznagel, who’s wearing a Brooks Brothers striped polo shirt and boat shoes during lunch at a restaurant across the bay from his summer home.
Stripey polo shirt and boat shoes ... classay. I think the implication is meant to be that he goes out for lunch without putting his trousers on.
Update: Definite evidence of unsoundness:
Last year, Spitznagel bought a 200-acre (80-hectare) cherry orchard and farm near his summer home in Northport Point, Michigan, that he’s converting into a goat dairy. Driving through the orchard, he lowers a window in his Mercedes-Benz G 55 AMG sport utility vehicle, plucks a fresh cherry off one of his trees and pops it in his mouth.
"This is my idea of farming," he says.
JKG has a lovely little essay on the practice of rich people buying up abandoned farms (summary: try to resist the urge to farm it, if you really can't, timber). He has some fairly condign words for the people who try to innovate dairies in non-dairy territory.
this item posted by the management 10/08/2011 06:21:00 AM
Monday, October 03, 2011
Yes but no
I basically agree with this, but there's an important nuance that gets missed. Brad writes, in re foreign exchange intervention and monetary intervention generally:
You want to get the speculators on your side--you want them to think that there is money to be made by betting that your policy will be carried out, rather than that there is money to be made by betting that your policy will break down. Thus the costs of deciding to do whatever it takes and of standing ready to do whatever it takes and announcing that you are going to do whatever it takes will probably, probably be surprisingly low.
Yes, that you want to get the speculators on your side and use them as a force multiplier. But we need to get much tighter on "you want them to think that there is money to be made by betting that your policy will be carried out, rather than that there is money to be made by betting that your policy will break down". Buzzword for the weekend's cocktail parties is the concept of "leaning against a bid", "having a bid to lean against". Let me explain:
Case 1, which should not be done because it is wrong and illegal. Say I am a trader who handles client orders and also trades the firm's capital. I have no special information about the stock of XYZ plc; I don't care whether I am long or short. The price of XYZ is bid 101, offer 99. Now a client places a big buy order in XYZ, at market. Am I long or short?
Smart kids at the front say "I am long, because the client placing a buy order is possibly trading on information, so the order has information content". Let's assume otherwise. Say we know that the order has no information content; it simply happens because XYZ has just been included in an index.
I am still long, on risk/reward grounds. Because I know that if I am short and wrong, I am exposed to normal market risk; if XYZ strikes oil, then the price could gap up on me. But if I am long and wrong, then I know the price cannot gap down on me; if I change my mind I can just sell my long to the client, because he put an order at market. Obviously, if the client had placed a limit order with a limit of 100, then I would not be able to "lean against" this.
Don't do that, by the way - leaning against a client order like this is called "front running" and not only is it market abuse offence basically everywhere, it's also very obviously unethical because by putting your own order in ahead of the client, you're moving the market against them and thus not doing your duty of best efforts execution. Similar offence; putting on a position ahead of publishing a research note that you expect to generate a flood of client orders.
Sorry, where was I?: Oh yes. But anyway, what if it's not you order that's come in, but you have a savvy idea that there is a big client order out there? Like, to take a real world example, you might have seen that a large insurance company had got into regulatory trouble and needed to liquidate a big portfolio of equities. This is public knowledge, so it can't be illegal to deal on it, but if you know there's a big sell order in at market, are you long or are you short?
A fortiori, of course, if a central bank has specifically put out its bid in the market, for the purpose of leaning against it, what do you do? The important thing here is not the size of the order, but the fact that it's completely price insensitive. If you just know that there's a big order that someone would kind of like to execute, then you're subject to market risk; if you know that the order has to be executed, either at market or at a specific price, then you know that there is guaranteed liquidity on one side of the trade, and the risks and rewards are skewed. It wasn't the size of the SNB intervention that moved the market - it was the determination.
In general, economics would be a lot better served if there was much more crossover between macroeconomists and market microstructure theorists. (by the way, secret weapon! this book. Absolutely kills it on the subject of exchange rate economics - if Frankel & Froot (1990) is a Sopwith Camel, this is something close to a General Contact Unit. Buy it for the global macro hedge fund manager in your life today).
this item posted by the management 10/03/2011 02:45:00 PM
Always pushing up the wrong channel
Look, I want to believe in the credit channel, always have done (ever since I worked in macroprudential surveillance, as its existence would have made my division institutionally more important at the Bank of England than it actually was). But it doesn't, it really doesn't.
It really looks plausible that there might be an exogenous channel by which bank lending could contribute to investment demand. You can tell all sorts of sensible-sounding theoretical stories, which presumably account for the extraordinary tenacity of Bernanke & Blinder (1988). Basically, they are all variations of some sort or another on the generic Fatty-and-JimBob model set out below:
Generic Fatty/Jim-Bob Credit Channel Model
Fatty is a bank; he has capital, he borrows money in the deposit market or some other how, and he makes loans to the various community of Jim-Bobs, who all run small businesses. The Jim-Bobs form a continuous distribution over the rates of return of their projects - if they can cover their financing costs, then they will apply for a loan from Fatty. For some reason or other, Fatty does not lend money to all the Jim-Bobs who want loans (and whose rates of return exceed the social cost of capital). Possible reasons:
1) Fatty only has limited regulatory capital, and so he has to ration lending more or less at random
2) Fatty's cost of funds has a risk premium built into it, so he can't lend to Jim-Bobs who don't earn more than his own cost of funds (which is higher than the social optimum)
3) Various blah about asymmetric information.
All of these models end up with some version of - make Fatty's business more profitable, reduce his cost of funds, buy some of his Jim-Bob loans off him to free up capital, guarantee some of the JB loans, and Fatty will make more loans to more Jim-Bobs - which will have a similar effect on investment spending as would government spending. But so much cheaper! And largely off budget! And politically acceptable!
The trouble is, it doesn't work that way. Basically the problem is our old friend nonergodicity. Not only do Jim-Bobs not form a well-behaved and known distribution of their rates of return, they don't form a distribution at all. They're all individual speculative, uninsurable risks and the decision to lend to them or not is not taken at the margin (old London Business School proverb, taught in the context of estimating the correct cost of capital for use in the Net Present Value Rule - "If changing your cost of capital by one per cent makes a difference between positive and negative NV, rest assured - it's negative".
Basically, the near-equilibrium situation in which a movement in Fatty's cost of capital, or funding, or even some insurance against the assymetric information risks he faces makes a difference by shifting some marginal projects from positive to negative NPV, is never the one that pertains. Institutionally, banks are more like a light switch than a tap - they're either in full risk-on mode (in which case they will lend to any project that looks kindasorta OK, and deal with the funding and capital consequences later), or in risk-off, batten-hatches mode (in which case they will do anything to avoid putting more capital at risk, because any incentive, be it carrot or stick, that you might want to set out to "get the banks lending again", is less of a factor than the danger of losing the money".
The point here is that Keynes The Master did not fuck up by not including a separate banking industry in his model; the investment decisions and motivations of the bankers in deciding whether or not to provide finance to the rest of the economy are exactly the same as those of industrial capitalists and are equally adequately described by Chapter 15. Bankers make their decisions based on whether loans will succeed or fail, which depends on their assessment of demand in the economy, and the model closes. Small-ticket subsidies and buggering about with guarantee schemes, and securitisations and blah don't work, unless they are big enough to constitute a fiscal policy in and of themselves. And I think that we can all see why, if carrying out fiscal policy is your aim, then doing so via those institutions which are the British banking system is unlikely to be the most efficient way in which to do so.
To an extent, this is me and the Old Keynesians versus Bernanke, Brad and the New Keynesians. A liquidity trap doesn't come about because of an increase in risk aversion or a shift in the demand for risk-free assets or anything like that - for some purposes it's useful to model it that way, but that isn't how it happens. Any model which has a smooth function at its heart isn't really Keynesian, including most of those in the General Theory in chapters other than 15. People don't just suddenly wake up feeling a bit more nervy and highly strung - their "risk aversion" increases because they are worried more about some specific bad thing about to happen to things they were previously thinking about investing in.
Which is why the latest proposal from the coalition to revive the good old small business loan guarantee scheme (with added bells, balls and whistles to make it look more like a CDO or like Fannie Mae, help us all) are going nowhere. These things only ever work at all, even in the dull-but-worthy, largely-economically-irrelevant sense as a form of industrial and/or regional policy, and the fact that they basically represent industrial policy gives you a clue that they're going to be done in way too small size. If we happen to get out of this crisis, it won't be because of that.
this item posted by the management 10/03/2011 01:37:00 PM
|