Third wave­func­tion points out that grad­u­ate exams are actu­ally tests that you con­form to cer­tain social stan­dards.:

1. Does the can­di­date require sleep?…
2. Does the can­di­date have any respon­si­bil­i­ties out­side of her career that can’t be put aside for three days straight? (e.g. tak­ing care of a child or other rel­a­tive, other house­hold responsibilities)
3. Does the can­di­date have some­one who will take care of her, bring­ing her food and wash­ing the dishes, tol­er­at­ing her com­pletely ignor­ing him and start­ing to smell from lack of hygiene?
4. Does the can­di­date require food on a reg­u­lar basis, or to admin­is­ter med­i­cine, or any­thing else that will take time away from work?

Over the last two weeks I’ve been try­ing to save my aca­d­e­mic career. Not from the Acad­emy itself, who seem rea­son­ably friendly (in aggre­gate) and will­ing to bear some of my inces­sant skep­ti­cal eyebrow-​​raising. But rather from the forces of Life: the increas­ing pres­sure to be mak­ing an actual liv­ing, the devel­op­ment of trou­bling fam­ily health mat­ters, my ridicu­lous belief that I should be work­ing no more than 40 hours a week on home­work and teach­ing oblig­a­tions, my inevitable inter­est in things that are Not Part Of The Thesis.

From me, in other words.

The Life Aca­d­e­mic is not a thing you can do in par­al­lel with… well, any­thing at all. You’re a fool if you think you can have a fam­ily and a house and a Pay­ing Job and grad­u­ate school all at once. You might be able to swing it, but the odds are stacked against you. Even pro­fes­sors need nan­nies and honorariums.

So in order to stave off the next wave of attack from the direc­tion of Real Life, I’ve been try­ing to Rethink It All. This involves reac­ti­vat­ing my social net­work, which has been lan­guish­ing while I’ve been in-​​country at the U, and call­ing in some social cap­i­tal chits, and gen­er­ally hav­ing Lots of Lunches. Meet­ing with friends, and col­leagues, and peo­ple who might have money I could scrounge, and in the mean­time try­ing to build a new con­sul­tancy with col­leagues scat­tered quite lit­er­ally (and sur­pris­ingly evenly) around the world. And so in per­son and online I talk with every­body, and try to get enthused about the Big Thesis-​​y Pic­ture and explore alter­na­tive stuff and new ways of think­ing, and mak­ing a suf­fi­cient liv­ing, and mak­ing the world a bet­ter place.

For me. And other peo­ple, sure, but for my fam­ily and me above all else.

So lots of talk­ing. Name-​​dropping. You know the per­son who has come up most often in this?

You would think it might be some famous per­son or hub­like social net­worky fel­low, some­body we (my social net­work) All Know. Mark. Or Cosma. Or Howard Rhein­gold. Or some Uni­ver­sity per­son (which Uni­ver­sity is after all the Biggest Log in my Log­jam of Life) like my aca­d­e­mic advi­sor or the Dean of the School of Infor­ma­tion (why she and SI are involved in the net­work is harder to explain here than elide) or maybe Stu Kauff­man, my old aca­d­e­mic advi­sor from Another Life.

You’d think that.

The per­son who has come up more often than any other in my innu­mer­able exhaust­ing con­ver­sa­tions of the last two weeks?

John Dewey.

Michael Cohen (who is an old colleague-​​of-​​colleagues, and who I chat­ted with the other day about a joint Ph.D. degree in my IOE Depart­ment and the School of Infor­ma­tion) said he was explor­ing Dewey’s sur­pris­ingly unre­marked role in the­o­ries of organizations—a role sup­planted to some extent by Herb Simon when he came along. Dewey and his stress of the impor­tance of habits in indi­vid­ual and group behav­ior came up sev­eral times.

Erik Schultes (who is an old friend from the Santa Fe Insti­tute and class­mate at the 1991 Com­plex Sys­tems Sum­mer School) is now a film­maker here in Michi­gan try­ing to make an ambi­tious project doc­u­ment­ing the Octo­pus via mod­ern net­works the­ory, and as we walked away from each other he asked us where Dewey’s house is. Down the street, as it happens.

Cosma men­tions Dewey, or maybe I men­tion Dewey to Cosma, and says he’d like to attend more to the great man’s work.

And finally, a con­stant. Sit­ting on the game table under the win­dowsill in the din­ing room, a stack of local Ann Arbor news­pa­pers from the 19^th Cen­tury. I sal­vaged them months ago, in hopes of dig­i­tiz­ing them, but the tech­nol­ogy for news­pa­per scan­ning is pretty shabby these days. In the top­most vol­ume, of the Ann Arbor Demo­c­rat, an amus­ing satir­i­cal arti­cle about some events we will prob­a­bly never under­stand. Some local ker­fluffle, some bawdy mob, some drunken stu­dent prank. And who do they men­tion by name? They say “The Admin­is­tra­tion”, but they name “Pro­fes­sor Dewey.”

Per­haps, given a run like this, so should we all. Men­tion him. Read him. Sur­pris­ingly, very few of his works have made it into Project Guten­berg by way of Dis­trib­uted Proof­read­ers.

I think I’ll do what I can to fix that in the com­ing days.

Actu­ally, not demo­graph­ics so much as cul­tural dynamics.

How has it come to pass that I can name two or three hand­fuls of peo­ple in my cohort (35−45) who started off as lab mol­e­c­u­lar biol­o­gists in the early 90s, and ended up say­ing “fuck this,” and wan­der­ing off into some other line of work?

And at the same time, I can name two or three hand­fuls of peo­ple younger than that, who are not lab mol­e­c­u­lar biol­o­gists but who now want to become one, or at least “get in on this bio­engi­neer­ing and structural/​systems biol­ogy stuff?”

But the ques­tion is not why there are two groups. The ques­tion is, assum­ing they rep­re­sent a sam­ple of a larger cul­tural dynam­i­cal process, will this accel­er­ate the attri­tion of the gen­er­a­tion of fac­ulty who cre­ated the first group, and who are unaware of the sec­ond group?

Please?

At Big Mon­key, Helpy Chalk, go read some cal­cu­la­tions of worth in “The World of Adjuncts”.

Just what crossed my desk this morn­ing. No rec­om­men­da­tion for, nor argu­ment against; just sayin’ I’m lookin’ at ‘em.

Abstract: We com­pared fore­casts of stock mar­ket volatil­ity based on real-​​time and revised macro­eco­nomic data. To this end, we used a new dataset on monthly real-​​time macro­eco­nomic vari­ables for Ger­many. The dataset cov­ers the period 1994–2005. We used a sta­tis­ti­cal, a utility-​​based, and an options-​​based cri­te­rion to eval­u­ate volatil­ity fore­casts. Our main result is that the sta­tis­ti­cal and eco­nomic value of volatil­ity fore­casts based on real-​​time data is com­pa­ra­ble to the value of fore­casts based on revised macro­eco­nomic data.

Abstract: This paper stud­ies the behav­iour of Inter­net prices. It com­pares price rigidi­ties on the Inter­net and in tra­di­tional brick-​​and-​​mortar stores and pro­vides a cross-​​country per­spec­tive. The data set cov­ers a broad range of items typ­i­cally sold over the Inter​net​.It includes more than 5 mil­lion daily price quotes down­loaded from price com­par­i­son web sites in France, Ger­many, Italy, the UK and the US. The fol­low­ing results emerge from our analy­sis. First, and con­trary to the recent find­ings for com­mon CPI data, Inter­net prices in the EU coun­tries do not change less often than online prices in the US. Sec­ond, prices on the Inter­net are not nec­es­sar­ily more flex­i­ble than prices in tra­di­tional brick-​​and-​​mortar stores. Third, there is sub­stan­tial het­ero­gene­ity in the fre­quency of price change across shop types and prod­uct cat­e­gories. Fourth, the aver­age price change on the Inter­net is rel­a­tively large, but smaller than the respec­tive val­ues reported for CPI data. Finally, panel logit esti­mates sug­gest that the like­li­hood of observ­ing a price change is a func­tion of both state– and time-​​dependent factors.

Abstract: This paper explores the deter­mi­nants of cor­po­rate fail­ure and the pric­ing of finan­cially dis­tressed stocks using US data over the period 1963 to 2003. Firms with higher lever­age, lower prof­itabil­ity, lower mar­ket cap­i­tal­iza­tion, lower past stock returns, more volatile past stock returns, lower cash hold­ings, higher market-​​book ratios, and lower prices per share are more likely to file for bank­ruptcy, be delisted, or receive a D rat­ing. When pre­dict­ing fail­ure at longer hori­zons, the most per­sis­tent firm char­ac­ter­is­tics, mar­ket cap­i­tal­iza­tion, the market-​​book ratio, and equity volatil­ity become rel­a­tively more sig­nif­i­cant. Our model cap­tures much of the time vari­a­tion in the aggre­gate fail­ure rate. Since 1981, finan­cially dis­tressed stocks have deliv­ered anom­alously low returns. They have lower returns but much higher stan­dard devi­a­tions, mar­ket betas, and load­ings on value and small-​​cap risk fac­tors than stocks with a low risk of fail­ure. These pat­terns hold in all size quin­tiles but are par­tic­u­larly strong in smaller stocks. They are incon­sis­tent with the con­jec­ture that the value and size effects are com­pen­sa­tion for the risk of finan­cial distress.

Abstract: A very promis­ing lit­er­a­ture has been recently devoted to the mod­el­ing of ultra-​​high-​​frequency (UHF) data. Our first aim is to develop an empir­i­cal appli­ca­tion of Autore­gres­sive Con­di­tional Dura­tion GARCH mod­els and the real­ized volatil­ity to fore­cast future volatil­i­ties on irreg­u­larly spaced data. We also com­pare the out sam­ple per­for­mances of ACD GARCH mod­els with the real­ized volatil­ity method. We pro­pose a pro­ce­dure to take into account the time defor­ma­tion and show how to use these mod­els for com­put­ing daily VaR.

Abstract: Ratio­nal herd behav­ior and infor­ma­tion­ally effi­cient secu­rity prices have long been con­sid­ered to be mutu­ally exclu­sive but for excep­tional cases. In this paper we describe con­di­tions on the under­ly­ing infor­ma­tion struc­ture that are nec­es­sary and suf­fi­cient for infor­ma­tional herd­ing. Employ­ing a stan­dard sequen­tial secu­rity trad­ing model, we argue that peo­ple may be sub­ject to herd­ing if and only if there is suf­fi­cient amount of noise and, loosely, their infor­ma­tion leads them to believe that extreme out­comes are more likely than mod­er­ate ones. We then show that herd­ing has a sig­nif­i­cant effect on prices: prices can move sub­stan­tially dur­ing herd­ing and they become more volatile than if there were no herd­ing. Fur­ther­more, herd­ing can be per­sis­tent and can affect the process of learn­ing. We also char­ac­ter­ize con­di­tions for con­trar­ian behav­ior. Our analy­sis sug­gests that herd­ing (and con­trar­ian behav­ior) may be more per­va­sive than was orig­i­nally thought. Hence, the paper pro­vides a new per­spec­tive on herd­ing in finan­cial mar­kets with effi­cient prices.

Abstract: What fac­tors deter­mine how well con­sumers make their actual choices with regard to finan­cial prod­ucts? This paper empir­i­cally eval­u­ates two dif­fer­ent choices con­sumers make when buy­ing deferred annu­ities. One choice con­cerns the type of insur­ance pol­icy, the other con­cerns the choice of insur­ance provider. For both choices we will analyse what fac­tors explain the qual­ity of the choice made. In par­tic­u­lar, we will inves­ti­gate the role of finan­cial advice in the deci­sion mak­ing process. By com­bin­ing Dutch con­sumer sur­vey data and data on quo­ta­tions by Dutch life insur­ance com­pa­nies, we obtain the fol­low­ing results. First, respon­dents who buy their pol­icy directly from an insurer attain a sig­nif­i­cantly bet­ter match between their risk pref­er­ences and the type of pol­icy cho­sen than respon­dents who pur­chase their pol­icy through an insur­ance bro­ker. Sec­ond, respon­dents who buy their pol­icy through an insur­ance bro­ker obtain a sig­nif­i­cantly lower pay-​​out than respon­dents who pur­chased their pol­icy directly from an insur­ance com­pany. These results raise doubts about the func­tion­ing of both the mar­ket for finan­cial advice and the mar­ket for life insurances.

Abstract: Since the under­ly­ing of the weather deriv­a­tives is not a traded asset, these con­tracts can­not be eval­u­ated by the tra­di­tional finan­cial the­ory. Cao and Wei (2004) price them by using the consumption-​​based asset pric­ing model of Lucas (1978) and by assum­ing dif­fer­ent val­ues for the con­stant rel­a­tive risk aver­sion coef­fi­cient. Instead of tak­ing this coef­fi­cient as given, we sug­gest in this paper to esti­mate it by using the con­sump­tion data and the quo­ta­tions of one of the most trans­acted weather con­tracts which is the New York weather futures on the Chicago Mer­can­tile Exchange (CME). We will apply the well-​​known gen­er­al­ized method of moments (GMM) intro­duced by Hansen (1982) to esti­mate it as well as the sim­u­lated method of moments (SMM) attrib­uted to Lee and Ingram (1991) and Duffie and Sin­gle­ton (1993). This last method is stud­ied since we think that it can give sat­is­fac­tory results in the case of the weather deriv­a­tives for which the prices are sim­u­lated. We find that the esti­mated coef­fi­cient from the SMM approach must have improb­a­bly high val­ues in order to have the cal­cu­lated weather futures prices match­ing the obser­va­tions. This find­ing is in accor­dance with the results of the prior works which have shown the empir­i­cal fail­ures of the consumption-​​based asset pric­ing model.