Serving the Quantitative Finance Community

 
User avatar
mj
Posts: 12
Joined: December 20th, 2001, 12:32 pm

Physics prof. or Fin prof.

July 14th, 2008, 3:53 am

Well, i know two people with 2 phds fairly well. So it's not that rare. One has pure maths and finance, the other pure maths and economics. It's more like 4 years for the first one and 3 for the second.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 14th, 2008, 7:54 am

Quote Well, i know two people with 2 phds fairly well. So it's not that rare. One has pure maths and finance, the other pure maths and economics. It's more like 4 years for the first one and 3 for the second.Some universities in the UK allow you to submit for a PhD based on 4 or 5 good publications, so potentially a prof with around 100 good publications in his career would probably have made around 20 PhDs in his area of study. However, it would be more challenging IMO to have multiple PhDs from different hard sciences, i.e. PhD Physics, PhD Statistics, PhD Maths, PhD Chemistry, PhD Biology. I don’t think the number of people with that combination would be much more than 1 or 2. Stats Guy
 
User avatar
DominicConnor
Posts: 41
Joined: July 14th, 2002, 3:00 am

Physics prof. or Fin prof.

July 17th, 2008, 7:42 am

Although my intuition is that a "hard science" PhD is "harder" than a Finance PhD, I do not see this as a scientific statement.Have we taken randomly selected people of comparable talent, and observed who succeeds or fails ?No we have not.Also, it is the case that there is no generally accepted standard for awarding PhDs, I even know people who engaged in PhD in Media Studies.I do not believe any subject is inherently "easier" than any other, if you are genuinely trying to do something new, which is supposed to be a key aspect of a PhD.For instance, even in media studies there is intensely analytical work to be done, some of which would be very lucrative.There is optimisation of portfolios in the creation of a major film. Which is optimal :$ 20M for Keanu Reevesor$10M in extra special FX and $10 M more in marketing ?If you could reliably predict TV show audiences, from inputs available at the time of the pilot episode, you would make so much money.On the other hand, the stuff I put in the guide about some PhDs in physics being essentially the role of maintenance mechanic for a strange piece of equipment is drawn from the experiences of many I talk to.But as a pimp, my advice is to do "hard science" because that is easy for me to sell, assuming everything else to be equal.I must add a disclaimer here of course. My evidence is historical, but if you are choosing a PhD now, you are interested in the labour market of 5 years from now, and on from that.My intuition is that the gap between the career utility of physical sciences vs finance will decline, but remain for many years to come.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 17th, 2008, 9:39 am

QuoteHave we taken randomly selected people of comparable talent, and observed who succeeds or fails ?This is a hard thing to do as everyone is different (take out twins, triplets etc..). It takes different skills to be good in different areas, some more people based and other more technically based etc.. So when we say comparable talent that will be biased on the opinions of the person carrying the analysis as at some stage they have to define how you group similar talents together which is highly subjective. QuoteBut as a pimp, my advice is to do "hard science" because that is easy for me to sell, assuming everything else to be equal.Which areas of science would you classify as a hard science? I would say Maths, Stats, Physics and perhaps Chemistry ?Given you must meet a wider pool of PhDs I was wondering if you have ever met someone with PhD Physics, PhD Statistics, PhD Maths, PhD Chemistry and PhD Biology ?Stats Guy
Last edited by StatGuy on July 16th, 2008, 10:00 pm, edited 1 time in total.
 
User avatar
twofish
Posts: 0
Joined: February 18th, 2005, 6:51 pm

Physics prof. or Fin prof.

July 17th, 2008, 1:11 pm

QuoteOriginally posted by: JamesHHWhat if someone writes a quality thesis without going to grad. school? Can't they submit it and get the PhD degree?If you've written a quality thesis then you've gone to grad school. It may not be formal grad school, but it's still grad school. There is this thing called the British Research Ph.D. in which you can convert a thesis into a credential, but even this takes a few years. Getting a Ph.D.'s is a lot like consecration of bishops. All you really have to do to get a Ph.D. is to have five people respected in the field get together and say that you deserve a Ph.D., and you have a Ph.D.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 17th, 2008, 2:08 pm

QuoteGetting a Ph.D.'s is a lot like consecration of bishops. All you really have to do to get a Ph.D. is to have five people respected in the field get together and say that you deserve a Ph.D., and you have a Ph.D. The danger of having 5 top people in the area reviewing your PhD is that they will pick holes in it left right and centre. There are always ways of improving a PhD (this is the further research section of the thesis, i.e. the bits you couldn't do and leave it to the next generation of geeks) I have been to some academic conferences in statistics and even the best in the field get run down after proposing some very good ideas. So you may devise a new techniqne and think you the next best thing, but someone who has been in the area for 30 plus years will say something like "the wider class of models from this approach is restrictive as opposed to z back in 1980 etc.." So some can really put you down and show you that you really don't know much. Outside academia a PhD is seen as a high academic achievement (since there is some ignorance about what actually goes on in academic from those outside), but inside it's only the starting point (or near bottom of the pile).Stats Guy
Last edited by StatGuy on July 16th, 2008, 10:00 pm, edited 1 time in total.
 
User avatar
aiQUANT
Posts: 0
Joined: June 4th, 2008, 6:20 pm

Physics prof. or Fin prof.

July 18th, 2008, 8:41 am

QuoteOriginally posted by: StatGuyI have been to some academic conferences in statistics and even the best in the field get run down after proposing some very good ideas. So you may devise a new techniqne and think you the next best thing, but someone who has been in the area for 30 plus years will say something like "the wider class of models from this approach is restrictive as opposed to z back in 1980 etc.." So some can really put you down and show you that you really don't know much. I get the feeling this happens more in the so called hard sciences like pure maths where people just faff about trying to make their work look difficult without solving real a problem. Naturally these people will try to diss each other in an attempt to feel important. Besides that progression in the Mathamatical domain is just not progression if viewed from an Engineering standpoint. Its just ludicrous.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 18th, 2008, 9:31 am

QuoteI get the feeling this happens more in the so called hard sciences like pure maths where people just faff about trying to make their work look difficult without solving real a problem. Naturally these people will try to diss each other in an attempt to feel important. Besides that progression in the Mathamatical domain is just not progression if viewed from an Engineering standpoint. Its just ludicrous.Statisticians at research level tackle some very tough applied problems from medical areas to finance. However, as with most applied areas we can only do our best when it comes to modelling, and someone who have a good maths background coupled with statistics and has been in the area many years will pick holes in the generalization of the model class you used to solve the problem. I agree it can see rather ludicrous from an applied persons perspective, but for a pure person they probably have a deeper understanding of the techniques and therefore can generalize the results more, which is seen as being harder than applying some wrong model to real data which is always an approximation anyway, and in some people’s eyes easier. Stats Guy
Last edited by StatGuy on July 17th, 2008, 10:00 pm, edited 1 time in total.
 
User avatar
aiQUANT
Posts: 0
Joined: June 4th, 2008, 6:20 pm

Physics prof. or Fin prof.

July 18th, 2008, 10:05 am

QuoteStatisticians at research level tackle some very tough applied problems from medical areas to finance. However, as with most applied areas we can only do our best when it comes to modelling, and someone who have a good maths background coupled with statistics and has been in the area many years will pick holes in the generalization of the model class you used to solve the problem. I agree it can see rather ludicrous from an applied persons perspective, but for a pure person they probably have a deeper understanding of the techniques and therefore can generalize the results more, which is seen as being harder than applying some wrong model to real data which is always an approximation anyway, and in some people’s eyes easier. Thats where the problem is. Pure people like generalizing more than necessary. From my experience it's usually the case that the applied person has a more relevant understanding of the theory in-order to solve the problem. The pure person easily gets lost in non-sensical generalizations whereas the applied person tries to keep things particular to the problem. For the pure person the bounds are not clearly defined because they don't know what they are trying to solve.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 18th, 2008, 10:46 am

QuoteThe pure person easily gets lost in non-sensical generalizations whereas the applied person tries to keep things particular to the problem. For the pure person the bounds are not clearly defined because they don't know what they are trying to solve.Yes I agree, but I am an applied stats person. From the statistical publications in top journals I have seen, some of the better articles (i.e. ones where the feedback from top statisticians in the world have been positive) are usually applied to an area x and then generalized and therefore can be applied to a wider set of areas. This is very very tough to do in a research area, and hence its viewed as harder since looking at the bigger picture in terms of model development is important to someIf you look at most of the hard applied science like physics, engineering etc.. they all analyze mass amounts of data and fit some fancy statistical models etc... but since they don't get the opportunity to meet a world leading statistician they usually get away with a restrictive class of statistical model as you usually only get to do the cutting edge statistical developments in statistical research. Data analysis is a specialized area, just like physics and the other hard science areas. So only with years of experience doing data analysis, ideally as a statistician, you get to understand the deeper problems.Stast Guy
Last edited by StatGuy on July 17th, 2008, 10:00 pm, edited 1 time in total.
 
User avatar
aiQUANT
Posts: 0
Joined: June 4th, 2008, 6:20 pm

Physics prof. or Fin prof.

July 18th, 2008, 11:12 am

QuoteOriginally posted by: StatGuyYes I agree, but I am an applied stats person. From the statistical publications in top journals I have seen, some of the better articles (i.e. ones where the feedback from top statisticians in the world have been positive) are usually applied to an area x and then generalized and therefore can be applied to a wider set of areas. This is very very tough to do in a research area, and hence its viewed as harder since looking at the bigger picture in terms of model development is important to someIf you look at most of the hard applied science like physics, engineering etc.. they all analyze mass amounts of data and fit some fancy statistical models etc... but since they don't get the opportunity to meet a world leading statistician they usually get away with a restrictive class of statistical model as you usually only get to do the cutting edge statistical developments in statistical research. Data analysis is a specialized area, just like physics and the other hard science areas. So only with years of experience doing data analysis, ideally as a statistician, you get to understand the deeper problems.Stast GuyI know you are an applied person like myself. I was referring to people like TJ who have done a pure science.The generalization thing might be necessary for statistics because the techniques need to be applied to other domains. So if they don't work generally then the research is useless. Statistics is still closer to pure science than applied science because of the general requirement that the methods should work for other problems. I'm not saying it's a bad thing. It's just an observation.
 
User avatar
twiceasnice
Posts: 0
Joined: July 8th, 2008, 10:07 pm

Physics prof. or Fin prof.

July 18th, 2008, 11:15 am

There is an american child prodigy doing 3 hard science PhDs simultaneously.
 
User avatar
StatGuy
Posts: 0
Joined: November 20th, 2007, 9:03 am

Physics prof. or Fin prof.

July 18th, 2008, 11:23 am

QuoteThere is an american child prodigy doing 3 hard science PhDs simultaneously.Yes but I don’t think there is 1 person in the world who has a PhD maths, PhD Statistics, PhD Physics, PhD biology and PHD chemistry. It would require a shift in thinking jumping from one hard science to another hard science coupled with the dedication required to finish one PHD, I just can’t see anyone who has that combination. But please let me know otherwise Quote Statistics is still closer to pure science than applied science because of the general requirement that the methods should work for other problems.To get a good Statistics PhD you need to be in the middle of applied and theory IMO. So you have some problem in domain x and you develop some statistical model to help solve that problem and if you can generalize that model to a wider class of models or show that it has the potential to be generalized etc. that would certainly enhance your PhD.Stats Guy
Last edited by StatGuy on July 17th, 2008, 10:00 pm, edited 1 time in total.
 
User avatar
ppauper
Posts: 11729
Joined: November 15th, 2001, 1:29 pm

Physics prof. or Fin prof.

July 18th, 2008, 12:32 pm

QuoteOriginally posted by: twiceasniceThere is an american child prodigy doing 3 hard science PhDs simultaneously.which would depend on the rules of the school:you need to take X courses as well as write a thesis (and pass prelims/qualifiers and language requirements) to get a PhD.At my school, you could not count the X courses taken for a PhD towards another PhD.Meaning the prodigy would have to take 3X courses
 
User avatar
TraderJoe
Posts: 1
Joined: February 1st, 2005, 11:21 pm

Physics prof. or Fin prof.

July 19th, 2008, 11:33 pm

QuoteOriginally posted by: StatGuyQuoteI get the feeling this happens more in the so called hard sciences like pure maths where people just faff about trying to make their work look difficult without solving real a problem. Naturally these people will try to diss each other in an attempt to feel important. Besides that progression in the Mathamatical domain is just not progression if viewed from an Engineering standpoint. Its just ludicrous.Statisticians at research level tackle some very tough applied problems from medical areas to finance. However, as with most applied areas we can only do our best when it comes to modelling, and someone who have a good maths background coupled with statistics and has been in the area many years will pick holes in the generalization of the model class you used to solve the problem. I agree it can see rather ludicrous from an applied persons perspective, but for a pure person they probably have a deeper understanding of the techniques and therefore can generalize the results more, which is seen as being harder than applying some wrong model to real data which is always an approximation anyway, and in some people’s eyes easier. Stats GuyWhat's the topic of your PhD ?