Do you guys find that agency theory and information economics helps you do your trading/structuring/quant jobs and improve your quant models? If so, in what sense?

QuoteOriginally posted by: KackToodlesDo you guys find that agency theory and information economics helps you do your trading/structuring/quant jobs and improve your quant models? If so, in what sense?What's it worth?

221 page views and still now answer? anybody? Let's air out some pro/con rants.

Make that 678. Popular thread .

OK, I'll take a timid step into the vacuum of this thread with three comments:1. Agency theory underpins some aspects of behavioral finance in explaining the actions of fund managers. The structure of the incentives for a fund manager mean that they don't always maximize returns or alpha. Maximizing their cut of the returns and minimizing the appearance of "bad decisions" changes how fund managers trade. 2. Warren Buffet's explicit avoidance of IPOs is also explicable from agency theory in terms of the company manager's choice of timing of the IPO, hidden information about the company, and the incentives for the company. Buying a share of stock is an implicit contract for the presumed future returns on the investment.3. At some level, agency theory and information economics are not compatible with some market theories because the former implies bounded rationality and incomplete information while the latter tends to assume perfect rationality and complete information. Are there any versions of CAPM that factor in the cost of information or the potential for share creators to game the system with judiciously planned secondary offerings or share buybacks?

QuoteOriginally posted by: Traden4AlphaAre there any versions of CAPM that factor in the cost of information or the potential for share creators to game the system ? Well, capm is all about what is observable in the data. Maybe all that info economics stuff is "philosophical" and cannot be discerned enough in the data to make it worthwhile to model in capm/apt framework.

Information economics is philosophical in the sense of the roots of the word philosophical (phil = liking & soph = complexity). Information economics can help us understand one of the complexities of the world -- how the non-zero cost of information affects the actions of producers and consumers. I would argue that it applies to financial markets in interesting ways. For example isn't declaring a dividend an example of signalling and isn't a conglomerate an example of bundling? Nor would I say that information economics can't be discerned in the data. Doesn't the fact that many investors under-diversify suggest some sort of cost-of-information effects? Isn't it clear that CAPM doesn't explain the data very well and that we either need to reject CAPM or find those add-ons that make CAPM more realistic?I suspect that two issues are the root of the problem. First, the proponents of CAPM would assert that information costs are negligible or are somehow averaged out or premiumed-out of the equation. At first glance, they may be correct because so many people can so easily find so much free information on the internet. At another level, they are terribly wrong because the true cost of information is dominated by interpretation labor, not access costs. CAPM assumes that the invisible hand has an all-seeing eye. Information economics recognizes that each participant's eye takes time and effort to observe the data before dipping their hand in the market.Second, one of the sad facts of academic economics is that so much of it is constrained by a decades-old edict to use analytic math. As wonderful as math is, it often requires ludicrous assumptions to ensure tractability. Analytic math is like a telescope in that it is great for zooming in on the logical structure of the universe but it is only useful in very limited set of circumstances (clear nights, distant objects, etc.).Perhaps there's a deep irony here in that the bounded capabilities of our smartest mathematicians require us to assume unbounded capabilities of our dumbest investors.

QuoteOriginally posted by: KackToodlesDo you guys find that agency theory and information economics helps you do your trading/structuring/quant jobs and improve your quant models? If so, in what sense?i avoid it like the plague.the market is made up of real participants and u need to understand their behaviour. u dont get that by reading theories - u learn it by doing it. just like u dont learn to play tennis by reading a book - especially one written by non tennis players.pw's approach is right - phenomenological, and to hell with dry theories.

fairplay I think that your last statement:"to hell with dry theories."may be intended to be a little tongue in cheek but nonetheless I have to protest. Without (dry) theory we would never graduate to practice. Most of academia could be accused of being 'dry theory' and yet most of practice has spawned from it. I am certain that theory and practice are inseparable.

QuoteOriginally posted by: Traden4AlphaI would argue that it applies to financial markets in interesting ways. For example isn't declaring a dividend an example of signalling and isn't a conglomerate an example of bundling? Agreed. I.E. speaks to "microstructure" and organizational effects. My question is whether I.E. has any relevance to finance, which is "meso-economics". The main themes of finance is diversiification and no arbitrage. Does I.E. surivive diversification? Does I.E. help to identify arbitrage opportunities? Unfortunately, it seems the answer to both questions is No and No. QuoteSecond, one of the sad facts of academic economics is that so much of it is constrained by a decades-old edict to use analytic math. As wonderful as math is, it often requires ludicrous assumptions to ensure tractability. Analytic math is like a telescope in that it is great for zooming in on the logical structure of the universe but it is only useful in very limited set of circumstances (clear nights, distant objects, etc.). Perhaps there's a deep irony here in that the bounded capabilities of our smartest mathematicians require us to assume unbounded capabilities of our dumbest investors. Why does this unfortunate math envy persists when it does not help economists arrive at more interesting results?

QuoteOriginally posted by: Traden4AlphaInformation economics is philosophical in the sense of the roots of the word philosophical (phil = liking & soph = complexity). Information economics can help us understand one of the complexities of the world -- how the non-zero cost of information affects the actions of producers and consumers.The most useful bit of information economics I've found is that it helps to understand labor markets which in turn helps one get a job. The dynamics of the jobs interview, headhunters, etc. etc. all can be seen in terms of information economics. Why do headhunters exist and why do employers never post their real names on monster.com with details about the job. The missing piece of information theory is that information isn't fungible. A bit of information that is useful to A is not useful to B. Two bits of information put together is usually more than the sum of the parts. This makes the interface between information and markets which deal with fungible and additive quantities interesting. Once you get into markets, the messiness of IE and agency gets hidden behind a number.

twofish, wow! Well put:The missing piece of information theory is that information isn't fungible. A bit of information that is useful to A is not useful to B. Two bits of information put together is usually more than the sum of the parts. This makes the interface between information and markets which deal with fungible and additive quantities interesting. Once you get into markets, the messiness of IE and agency gets hidden behind a number.This also has very interesting implications in non-market games like poker.

QuoteOriginally posted by: BetaExoticBetsThe missing piece of information theory is that information isn't fungible. Gong! This is soo obviously wrong its not even funny.Counterexample 1: A has information. B pays A $1000. A gives B information. (How do you think National Enquiror gets hold of rumors from free lance reporters, or China gets US nuclear secrets?!)Counterexample 2: Mutual fund manager has superior analytical skills to process information. You invest money in mutual fund and pay fund manager 0.5% management fee. Fund manager uses his information to help you manage your money.Counterexample 3: You want to learn options pricing. Professor has information about how to price option. You pay tuition. Professor teaches you a course on options pricing.Counterexample ad nauseum: DCFC has information about hot prospect. Bank wants to hire hot prospect. Bank promises DCFC payment. DCFC hands over hot prospect for interviews.

QuoteOriginally posted by: KackToodlesQuoteOriginally posted by: BetaExoticBetsThe missing piece of information theory is that information isn't fungible. Gong! This is soo obviously wrong its not even funny.Counterexample 1: A has information. B pays A $1000. A gives B information. (How do you think National Enquiror gets hold of rumors from free lance reporters, or China gets US nuclear secrets?!)Counterexample 2: Mutual fund manager has superior analytical skills to process information. You invest money in mutual fund and pay fund manager 0.5% management fee. Fund manager uses his information to help you manage your money.Counterexample 3: You want to learn options pricing. Professor has information about how to price option. You pay tuition. Professor teaches you a course on options pricing.Counterexample ad nauseum: DCFC has information about hot prospect. Bank wants to hire hot prospect. Bank promises DCFC payment. DCFC hands over hot prospect for interviews. I think you have a strange definition of fungible going here. A's gossip info is not fungible because it is not interchangeable with the mutual fund manager's superior analytical skills to process information or DCFC's hot prospect (unless Britney Spears is about to be hired as a hot new hedge fund manager)

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: KackToodlesQuoteOriginally posted by: BetaExoticBetsThe missing piece of information theory is that information isn't fungible. Gong! This is soo obviously wrong its not even funny.Counterexample 1: A has information. B pays A $1000. A gives B information. (How do you think National Enquiror gets hold of rumors from free lance reporters, or China gets US nuclear secrets?!)Counterexample 2: Mutual fund manager has superior analytical skills to process information. You invest money in mutual fund and pay fund manager 0.5% management fee. Fund manager uses his information to help you manage your money.Counterexample 3: You want to learn options pricing. Professor has information about how to price option. You pay tuition. Professor teaches you a course on options pricing.Counterexample ad nauseum: DCFC has information about hot prospect. Bank wants to hire hot prospect. Bank promises DCFC payment. DCFC hands over hot prospect for interviews. I think you have a strange definition of fungible going here. A's gossip info is not fungible A sold his information to B. How is it not "fungible" when it can be sold for $1000 cold hard cash?