Transparency Liquidity
Felix Salmon is a very smart person who writes very well. Also he once invited me to an instant messenger debate that he posted on his high traffic blog. So I’d like to make only polite criticisms. However, I can’t write well so I will please translate the following to polite in your heads.
Salmon wrote “The CDS market is actually more transparent, with smaller bid-offer spreads”. That is, Salmon equates “the CDS market is actually more transparent”, and “CDSs are more liquid.” Liquid and transparent are not synonyms. Take the metaphors literally, and look at an old analog thermometer. You will find that mercury is liquid but not transparent and glass is transparent but not liquid.
Serious discussion after the jump.
In the sentence which I mock above, Salmon is criticizing an incorrect claim in a New York Times editorial. The claim is “A big part of the problem is that derivatives are traded as private one-on-one contracts. That means big profits for banks since clients can’t compare offerings.”
As noted by Salmon this is not true. It could be true, but, in fact, people know the current price of CDS on something*.
On another topic, Salmon wrote “But it doesn’t necessarily mean lower trading costs for the buy side: just ask anybody who tries to buy and sell bonds listed on the Luxembourg exchange.” I think it is clear who Salmon means by “anyone who tries to buy and sell bonds”. This would not be any firm that ever issued a bond nor does it mean any investor who ever bought a bond. He is thinking of people who try to profit from active trading strategies. “Anyone” means “any trader.” This is Salmon’s point of view. He talks to professional traders. Often he criticizes them, but he thinks about their problems and challenges.
Here, however, he is commenting on an editorial discussing public policy. Obviously Salmon doesn’t think the final aim of public policy is to make the world convenient for traders, but he assumes that making the world convenient for traders will lead to good economic outcomes. He definitely thinks that smart people trying to beat the market make the market work better.
This would make sense if one accepting a strong by semi strong form efficient markets hypothesis where prices are optimal given public information, private information can be obtained at a price, and prices are what they would be if informed traders had rational expectations. This is exactly the dominant assumption in the finance literature. It is also clearly nonsense. Salmon assumes that traders were rational.
dangerous risk-taking is actually a good thing, in financial markets. When people engage in risky behavior on Wall Street, they stand to lose a lot of money, but they know that they stand to lose a lot of money, and government doesn’t end up having to step in and bail them out. The big systemic problems happen when leveraged actors think that they’re not engaging in risky, speculative behavior
So Salmon asserts that dangerous risk taking is a good thing, because when people take risks, they know they are taking risks, except for the people who don’t know they are taking risks. Does Salmon assert that the claim “people always know when they are taking risks” is plainly obviously true or plainly obviously false ? He asserts both, with equal confidence and absence of evidence.
I think I can guess what he thinks. The traders with whom he talks are smart, so they don’t take risks without knowing it. The former CEOs of Lehman and AIG are dumb. The problem is that one can be very smart without having rational expectations.
In the same passage, he also notes a cause of big systemic problems and confidently asserts that it is the only cause of big systemic problems. One could as well argue that all market crashes involve the text “.com.”
To quote Salmon, “What a mess.”
It is very easy to see that outcomes are not what they would be if informed traders were rational. Basically cut out the theoretical middle man. Salmon assumes that high trading volume leads to efficient pricing. High trading volume is what he means by “liquidity” and, here, “transparency.” Trading volumes have changed enormously. High trading volume always comes with high price volatility. Price volatility is vastly greater than it should be (google “Shiller” and “excess volatility”) . It is plainly obvious that, in the real world, markets with a high volume of trade are less efficient (in the sense of the efficient markets hypothesis) than markets with a low volume of trade.
History shows that making markets convenient for traders, including really smart traders, reduces the usefulness of the signals markets send to the real economy. That’s why titans of finance who become treasury secretary like Tobin taxes(note the absence of the word *all* those titans are Nicholas Brady who publicly supported one and Robert Rubin who inquired as to how one could be implemented). It is possible that Rubin is clueless about finance, but that is not the way to bet.
Now Salmon writes many smart things in his post. To paraphrase Salmon “the problem is that it gets to the right destination by using the kind of rhetoric which makes it seem as though the only people who are unhappy about [demanding] proposed [politicall unmentionable] derivatives regulation are the people who don’t understand the derivatives market.”
*I note in this footnote that the problem is that there is a current price of a CDS written on something. There shouldn’t be. Given counterparty risk, there should only be a price of a CDS written by someone on something. Comparing prices and buying the cheapest CDS is a great way to guarantee that if the underlying sercuirty defaults so will the CDS writer. I take that seriously. Sure traders had plenty of data and low bid ask spreads. However, IIUC the data were collected under the totally false assumption that counter party risk was negligible. That’s insane. It is like assuming that a mortgage is a mortgage and it doesn’t matter if the debtor documented income or just claimed income. In each case, totally incorrect assumptions of homogeneity were made so that one could make a big huge data set. Everyone did this so they guaranteed that their assumptions would be false – if someone assumes high or medium quality someone else can make a profit producing low quality.
The desire to play with computers caused people to forget that garbage in means garbage out. If all financiers had been math phobic and computer phobic, the world would be a better place today.
People too deep in make this problem too complex. The underlying root problem is low marginal and low cap gains tax rates combined with zero anti-trust enforcement. No public good is served by short term transactions in which parties walk away with $10’s of millions of gains. Instead of a bunch of micro-managing of rules, we need to break up the big financial oligopolistic firms and their mega-corporate clients, and tax the hell out of massive paydays. Bankers who want to earn $500K/year for 20 years and get a pension from the bank will behave differently from those who want to walk out with $50M from some cash-out.
You have pointed everything out here; I am simply restating in my own terms. If you make CDS too transparent then there could theretically be a run on (or shorting of) the companies feeling the need to purchases this form of insurance –however meaningless it is, thanks AIG– but for Felix Salmon to lay the foundation of part of his argument with rationality and market efficiency for an inherently opaque and illiquid market is utter nonsense. This, along with rational expectations theory, belongs to the dustin of economic history.
***This is Salmon’s point of view. He talks to professional traders. Often he criticizes them, but he thinks about their problems and challenges.***
Exactly. When I read Salmon’s article yesterday, I was not impressed. Somehow Mr Salmon seems to have overlooked the problem that AIG wrote maybe 150 BILLION DOLLARS in CDSs they could not cover — with disasterous results. And there is no reason that some other company or six other companies might not be doing the same thing today.
I’m not sure that these people ought to be allowed to play using real assets. Maybe as a public service we can set up a huge fantasy roll playing game where Mr Salmon and professional traders can wheel and deal to their hearts content. … Using play money.
There’s roughly 250 million cars registered in the U.S. Allstate is the number 2 writer of automobile insurance with roughly 10% market share, meaning they cover 25 million cars. The average price of a used car is roughly $10,000. They have about $17 bn in shareholders’ equity and $20 bn of insurance reserves.
So if all of Allstate’s auto policy holders get into a giant car accident tomorrow, they would have a $250 bn liability, or 7x their equity and reserves they could not cover. State Farm has roughly double the market share of Allstate, and I suspect is similarly “thinly” capitalized.
What’s the likelihood of all of Allstate’s insureds getting into an automobile accident at the same time? Pretty low, I admit, and to the naked eye would round to zero, but is it zero?
M.Jed:
Conflating the issue and no where is it stated that driving a car is the same as selling CDS, which is actually what you are comparing.
The fact of the matter is you can get into far more trouble driving a car with or without insurance than you can selling CDS or the equivalent of not having car insurance the nefarious naked CDS. Indeed, reckless car driving can result in prison sentences, suspension or revocation of license, etc. How many people have gone to prison after the latest Wall Street boondoggle? And how many bonuses were granted in spite of Wall Street dragging the economy down?
Driving a car is highly regulated hence a lower possibility of the ultimate calamity you suggest. Any fool on Wall Street can write a CDS and sell it on the inside of his trench coat because of little regulation. So the probability of everyone insuraed by All State having an accident is rather slim. On the other hand, we know what CDS did to the economy.