Pages

3/30/13

Since hedge funds and a lot of trading jobs today are dominated by quants do the people with an econ/finance deree without or little programming knowledge have any potential in the HF/Trading industry?
A lot of long/short equity funds are also hiring quants from what i have been reading. Are the days where an Einhorn or an Ackman could make it big over?

I really need to make a decision whether I move in the PE or HF path. Looks like PE seems more welcoming. xD

Comments (135)

In reply to IlliniProgrammer
Best Response
3/22/13

IlliniProgrammer:
Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

I have said almost nothing negative on these forums for over two years now but this is just ridiculous. Your holier than thou rants about drinking PBR on the PATH train, living in a Jersey City apartment, and one day living off 40k a year on a farm in Iowa are probably a good counterbalance to the college kids who think Wall Street really is just dressing fancy, cocaine, and blow jobs from hookers in limousines. At the same time, finding some kind of superiority in the fact your wallet could turn a lump of coal into a Tiffany diamond is annoying, and the fact that you are in a MSF program makes you feel like you have a better understanding of allocating capital than people who have 50+ year track records doing things a "traditional way" is blatant narcissism.

There are lots of ways to make money. I acknowledge that the people at RenTech are brilliant, and they deserve to charge exorbitant fees. Despite this, I would trust my capital to Klarman, Einhorn, or Buffet in a heartbeat over any quant fund. The guys at LTCM were the absolute best, multiple Nobel prize winners, the supposed best risk management in the business, etc. Baupost and Berkshire and may never have years where they are up 50% going forward, but barring a nuclear disaster I'm positive I'll never wake up one morning and see BRK quoted at 0, or that Seth Klarman will have to write a letter to his investors saying that he blew up. Finance at the end of the day boils down into two really simple things, buying a piece of something or loaning money, with an incredible amount of human, industry specific, and macroeconomic factors converging. There are many attributes and risks that can't be modeled. How would you model the fact that Aubrey McLendon is a scumbag who secretly had favorable financing terms to invest in the very best wells after initial output data was collected but not yet published? How would you model that Lehman's disclosures and footnotes were shadier than financial statements of a Camden pawn shop? How does a computer recognize that a business has incredible pricing power because the customer base has a largely intangible attachment?

Do you really think that an engineer with degrees in Math/CS/Physics etc could a better private equity operator than Schwarzman or Kravis? Buying a controlling stake in companies often requires having to deal with executives that are very strong-willed and resistant to change. Is the quant super hero going to soothe the ego of the CEO, negotiate with the labor union, identify which customers to target, which ones to abandon, work with bankers to sell the thing in 5 years, etc. I'll bet every nickel I have that PE will not be dismantled and replaced by a bunch of guys with quant backgrounds. It is an incredibly people intensive business. I'd fucking love to hear some quant guy tell Carl Icahn he could do a better job of turning around a business than him. The guy probably can't even spell heteroskedasticity, but he was up 35% in 2011 and over 20% last year (these are off the top of my head and I was out awhile tonight).

Regarding hedge funds, there are room for quant guys and traditional shops that care about things like quality of management and their incentives, the moat of the business model, pricing power, and a logical explanation for favorable industry economics. If you think that "traditional" hedge funds operate base on the material taught in MBA classes you are sorely mistaken. I've never built a dividend discount model, calculated an F score, utilized Treynor-Black or Black Litterman, modeled an efficient frontier etc in my time at a hedge fund.

I hope you find a job you really like after grad school and make a ton of money figuring out how to use computers to predict the prices of esoteric financial instruments. You are 100%, unequivocally wrong in your quoted assessment though.

The WSO Advantage - Hedge Funds

Financial Modeling Training

IB Templates, M&A, LBO, Valuation.

Hedge Fund Interview Questions

10+ webinars & actual pitches.

Resume Help from HF Pros

Land More Interviews.

Find Your HF Mentor

Realistic HF Mock Interviews.

3/21/13

looks opposite to me :)

In reply to animalz
3/21/13

And why is that?

3/21/13

knowledgeispower:
Since hedge funds and a lot of trading jobs today are dominated by quants do the people with an econ/finance deree without or little programming knowledge have any potential in the HF/Trading industry?
A lot of long/short equity funds are also hiring quants from what i have been reading. Are the days where an Einhorn or an Ackman could make it big over?

I really need to make a decision whether I move in the PE or HF path. Looks like PE seems more welcoming. xD

Don't be so short-term focused. The quant field remains in growth mode while the other strategies have been stagnant/declining as they struggled with returns. There has been a shift recently (I think last year, a lot of quant funds underperformed), with value strategies getting much more attention. Funds allocation seems to follow a retrospective cycle, getting into whatever just had the best returns.

3/21/13

Also remember that during the crash, tons of hedge funds failed using quant/automated strategies precisely because those strategies only work within certain preset parameters, but when a black swan event takes place you still need a human being to be able to navigate nasty waters.

3/21/13

Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

In reply to IlliniProgrammer
3/21/13

Thanks for the input. I was wondering how long it would take for a person who cant write code at all to learn it up? I have a pretty strong math background. I still have 2 more yrs to graduate. Is it worth taking the effort to learn up programming or should I just stick to what I know and use it well?

3/21/13

And in the years to come will the MBA/CFA guy find work in the hedge fund industry or is it just going to be the Comp Sc/Math Phd guy?

In reply to IlliniProgrammer
3/21/13

IlliniProgrammer:
Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

I foresee a more quantitative focus in analyzing bigger data sets in PE. There are going to be more software vendors that offer business analytic software that utilizes machine learning. I doubt you will need to code C++; a user friendly GUI will likely make much of this type of analysis open to more people.

Isn't that the idea behind software like Palantir - to make high level analysis simple and intuitive?

Please don't quote Patrick Bateman.

In reply to IlliniProgrammer
3/21/13

IlliniProgrammer:
The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

You're kidding, right? Right?

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

In reply to IlliniProgrammer
3/21/13

IlliniProgrammer:
Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

You have great advice on saving money, but should probably stray from advising anyone on hedge funds, because you're spreading misinformation. Most small funds aren't the types of chop shops you described, nor do they necessarily hire MBAs. In fact, MBAs are typically looked down on at most HFs. Furthermore, most large shops aren't quant-focused. Even when it comes to "sophisticated" large shops, the hedging they do is actually rather simplistic.

No, computers aren't better than humans at investing. They're better at scalping, but that's about it. Outside of a few exceptions, most quant funds do NOT succeed. Unfortunately, I can't provide data on the success rate of quants vs. human-managed shops, but suffice it to say that out of the top 25 performing funds each year, the vast majority tend to be human-controlled.

3/21/13

there are plenty of l/s equity/credit guys employing fundamental analysis to generate returns. I don't know what you are talking about in terms of those guys having no place in the hf world.

3/21/13

knowledgeispower:
Since hedge funds and a lot of trading jobs today are dominated by quants do the people with an econ/finance deree without or little programming knowledge have any potential in the HF/Trading industry?
A lot of long/short equity funds are also hiring quants from what i have been reading. Are the days where an Einhorn or an Ackman could make it big over?

I really need to make a decision whether I move in the PE or HF path. Looks like PE seems more welcoming. xD

Having quantitative skills is definitely useful (if not required), but don't drink too much of the quant kool aid...

Computers are great at automating relatively simple tasks and executing them very quickly. That's basically what the high-frequency trading/statistical arbitrage firms do: they train software to do some simple arbitrage, using historical data.

But software is completely useless when it comes to interpreting new and complex information. Yes, some long/short equity funds are indeed hiring quants to help the analysts crunch numbers. But it doesn't mean that quants are making the investment decisions.

3/21/13

I really think the future is bright for a dynamic programmer/mathematician/econ&finance guy.

Look, the fact is that technology appears to follow an exponential growth function and if you want to be in the most competitive field then guess what....you need to be competitive.

Speaking as a math/econ&finance guy with only a small amount of programming guess what I'm doing now...Learning programming. It is a bit of an exaggeration to think that the world will go completely to straight HFT/Quant traders who have no knowledge of markets. I really think it is healthy arbitrage right now forcing markets to be much more accurate.

How bullshit is it to hear a finance guy saying a price will go up because the company normally trades at some multiple of its ebitda...That shit is dead. Those people were idiots. I want to hear there is a 65% probability that the price goes up by 10-15% using current macro/industry/company metrics. a computer calculates all this in polynomial time (CS people can correct me if im wrong). And given that probability and our risk tolerance this is the amount that should be put into play to keep risk levels steady.

Its a new time now, embrace the fact that the meritocracy of trading has become more efficient. Market corrections are faster and more accurate meaning the probability of huge swings should decline. All we have to worry about is the black swan events which should be more effectively modeled with jump diffusion models.

Our economic models are converging to match the models that traders price with and risk levels are being calculated more effectively .Chances are we laugh at the days that we used to endure days of 3% swings in the s&p.

3/21/13

Those two posts by IP and jktecon are the perfect, crystal-clear example of why smart analysts who look at the big picture will always be able to make money. Most nerds have so much hubris that they usually run full-speed into thick brick walls (do I really need to mention the gigantic failures of the smartest guys in the room again?)

3/21/13

No one answered my post. Why programming? I mean I certainly understand the applicability now and its use, but it is not a panacea for financial analysts. As data analysis evolves there will be more off the shelf tools to conduct high level analysis. It's like the early days of computers, why pull up a command prompt when you can click on an icon.

With the exponential curve of technology...do you think a generalist who has taken some programming courses will be able to construct something useful? Best of breed programmers will focus on their niche and join their work together.

It's like way back in the day when your barber was your surgeon/doctor...

Please don't quote Patrick Bateman.

3/21/13

Well, there's programming and programming. Knowing some VBA, or even some Python or R can be quite useful for many data analysis tasks, and it's not that hard.

Knowing C++, advanced algorithms and the like is overkill for most people working in finance.

3/21/13

Schumpeter seems to share my logic on the importance of programming for future business people. No one is saying you need to learn to program highly efficient algorithms in c++. And certainly no one called it a cure-all so IDK where you got those sentiments.

I am advocating that you understand the way a computer operates. Most importantly no human being in the 21st century should be unaware of a basic do,while / for loop. Even if the computer becomes completely able to do what you want it to because of better operating systems you will likely still need these basic understandings.

Knowledge of stacks, queues, lists, linked lists and the like will probably be reserved for hardened professionals as it is now, but as a society I do believe most people allow themselves to be handicapped by the computer.

The shear fact that you think you are competing against arguably the greatest tool mankind has ever devised is a problem. Understand the machine and you will see its beauty, It is an efficient communicator. Needless to say a financial analyst who has abilities with VBA/excel is an asset.

AND YOU WON'T KNOW WHAT YOU CAN DO UNTIL YOU TRY IT! We don't know what goes on in your office/what tools people use/how they look at it. Programming is not as hard as you are probably making it in your mind. The first 10 hours may be arduous but from there it becomes fun. Build slowly, it's like learning to write an essay, but keep this in mind. If you reach even a third grade literacy rate, you will be considered a genius.

3/21/13

Fair enough guys. That is my view of the role MBAs and the well connected but otherwise incompetent often play in hedge funds. At the larger funds there aren't that many MBAs, so the people claiming MBAs and ordinary people with good work ethics and some combination of eliteness and people skills are all over hedge funds must be at the smaller shops.

In reply to Schumpeter
3/21/13

Schumpeter:
Well, there's programming and programming. Knowing some VBA, or even some Python or R can be quite useful for many data analysis tasks, and it's not that hard.

Knowing C++, advanced algorithms and the like is overkill for most people working in finance.


I disagree, somewhat. For a typical financier, even one at an investment bank, it's overkill to expect them to know algorithms perfectly.

But the truth is that good algorithms can be worth millions in finance and on the buy side that gets reflected a bit better.

In reply to IlliniProgrammer
3/21/13

IlliniProgrammer:
The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

Most of these funds are not stat arb or HFT funds.
http://www.thehedgefundjournal.com/sites/default/f...

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

In reply to Schumpeter
3/21/13

Thanks for the input. So now the question is, if someone does not have a programing backgroud does he have to go for an MFE maybe? Or all he needs to do is lern up enough programming (self taught?) to understand what the "quant" in his firm doing or saying to him?

3/21/13
3/22/13

An MFE would be complete overkill and you most likely don't have the background if you are making these kinds of arguments, so it would not be worth it. Learn programming on your own.

First look at some cool initial programs that people have done that inspire you. Figure out easy problems you have that you think a computer could help you with (e.g. solutions of linear equations, checking dv01 calculations, value of a company using different ratios and keeping other things constant). Everything I listed there can be done with fairly straightforward algorithms that you could probably learn in the next 24 hours.

Look for inspiration first. You don't have to take our word for how useful programming is. Once you have the inspiration you are well on the road to becoming a programmer.

3/22/13

Oh and congratulations on looking into becoming a 21st century businessman.

In reply to jktecon
3/22/13

Haha.Thanks for the advice :)

And btw is programming going to be a prerequisite for an entry even in a long/short equity fund in the years to come? Or going by the glut of MFE grads will people with a Graham/Buffet style approach have an edge?

Noob question, I know. But I need as much iinput as I can. Thanks.

The WSO Advantage - Hedge Funds

Financial Modeling Training

IB Templates, M&A, LBO, Valuation.

Hedge Fund Interview Questions

10+ webinars & actual pitches.

Resume Help from HF Pros

Land More Interviews.

Find Your HF Mentor

Realistic HF Mock Interviews.

3/22/13

You have to find an edge. We are telling you to use the most useful tool in finance outside your brain (aka the computer) to help you in that endgame goal.

The good MFE grads will find their edge, the bad ones will be glorified IT professionals at best. The good finance guy who knows some econ math and programming will be an asset when he has ideas. The bad finance guy who went to school still thinking that Rambo is the quintessential trading personality will probably be a bank teller (or an investment banker if he went to harvard...same difference).

My opinion is that there is no way that an educated businessman ever evaluates the guy who can program (even if slightly inferior otherwise but quite close in terms of education and work ethic) lower than a straight finance jock. You still need personality though so don't ever think that programming is now the end all. I and most other people here, I believe, are advocating a balance. Learn to speak, how to make friends, a decent amount of math, some programming, the finance and the economics, and learn how to lose but still keep the determination to be a winner.

The 21st century businessman...He's like everything else in this new renaissance, look at the old, learn from it but do not complacently try to fit the old mold...your goal is to look like you've broken it completely while still actually adhering to its core doctrines(they haven't changed throughout human history).

In reply to Gray Fox
3/22/13

Gray Fox:
IlliniProgrammer:
Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

I have said almost nothing negative on these forums for over two years now but this is just ridiculous. Your holier than thou rants about drinking PBR on the PATH train, living in a Jersey City apartment, and one day living off 40k a year on a farm in Iowa are probably a good counterbalance to the college kids who think Wall Street really is just dressing fancy, cocaine, and blow jobs from hookers in limousines. At the same time, finding some kind of superiority in the fact your wallet could turn a lump of coal into a Tiffany diamond is annoying, and the fact that you are in a MSF program makes you feel like you have a better understanding of allocating capital than people who have 50+ year track records doing things a "traditional way" is blatant narcissism.

There are lots of ways to make money. I acknowledge that the people at RenTech are brilliant, and they deserve to charge exorbitant fees. Despite this, I would trust my capital to Klarman, Einhorn, or Buffet in a heartbeat over any quant fund. The guys at LTCM were the absolute best, multiple Nobel prize winners, the supposed best risk management in the business, etc. Baupost and Berkshire and may never have years where they are up 50% going forward, but barring a nuclear disaster I'm positive I'll never wake up one morning and see BRK quoted at 0, or that Seth Klarman will have to write a letter to his investors saying that he blew up. Finance at the end of the day boils down into two really simple things, buying a piece of something or loaning money, with an incredible amount of human, industry specific, and macroeconomic factors converging. There are many attributes and risks that can't be modeled. How would you model the fact that Aubrey McLendon is a scumbag who secretly had favorable financing terms to invest in the very best wells after initial output data was collected but not yet published? How would you model that Lehman's disclosures and footnotes were shadier than financial statements of a Camden pawn shop? How does a computer recognize that a business has incredible pricing power because the customer base has a largely intangible attachment?

Do you really think that an engineer with degrees in Math/CS/Physics etc could a better private equity operator than Schwarzman or Kravis? Buying a controlling stake in companies often requires having to deal with executives that are very strong-willed and resistant to change. Is the quant super hero going to soothe the ego of the CEO, negotiate with the labor union, identify which customers to target, which ones to abandon, work with bankers to sell the thing in 5 years, etc. I'll bet every nickel I have that PE will not be dismantled and replaced by a bunch of guys with quant backgrounds. It is an incredibly people intensive business. I'd fucking love to hear some quant guy tell Carl Icahn he could do a better job of turning around a business than him. The guy probably can't even spell heteroskedasticity, but he was up 35% in 2011 and over 20% last year (these are off the top of my head and I was out awhile tonight).

Regarding hedge funds, there are room for quant guys and traditional shops that care about things like quality of management and their incentives, the moat of the business model, pricing power, and a logical explanation for favorable industry economics. If you think that "traditional" hedge funds operate base on the material taught in MBA classes you are sorely mistaken. I've never built a dividend discount model, calculated an F score, utilized Treynor-Black or Black Litterman, modeled an efficient frontier etc in my time at a hedge fund.

I hope you find a job you really like after grad school and make a ton of money figuring out how to use computers to predict the prices of esoteric financial instruments. You are 100%, unequivocally wrong in your quoted assessment though.

\\

What a beauty.

In reply to jktecon
3/22/13

Really well said. Thanks for the motivation. Appreciate all the help :)

3/22/13

Look to global macro or activist HFs.

In reply to John Daggett
3/22/13
In reply to Gray Fox
3/22/13

Gray Fox:
IlliniProgrammer:
Depends on whether you work for a Renaissance, Fortress, Citadel, or DE Shaw, or if you work for one of the more boutique funds.

The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

The smaller shops tend to involve MBAs randomly stepping out the room for a few minutes with their cell phones, and walking back in with a bizarre announcement on how some biotech stock is the next big hot ticket, buying, and watching the firm get FDA approval for something three days later.

PE has another 5-10 years before this hits them too, but turnarounds really follow a pretty straightforward strategy- cut costs while preserving the effectiveness of R&D and the customer experience. This is something that, quite frankly, a quant with an engineering background should be doing- largely using machine learning algorithms- rather than an MBA trying to scrap out an L1 regression.

The bottom line is that except in a few cases, computers are now better than humans at investing. The hedge funds have replicated the finance undergrad/MBA and made a bunch of copies of him, and that is doing a lot of the short-term trading now. There is still room for normal people, but probably not in the same traditional roles. Welcome to Skynet.

I have said almost nothing negative on these forums for over two years now but this is just ridiculous. Your holier than thou rants about drinking PBR on the PATH train, living in a Jersey City apartment, and one day living off 40k a year on a farm in Iowa are probably a good counterbalance to the college kids who think Wall Street really is just dressing fancy, cocaine, and blow jobs from hookers in limousines. At the same time, finding some kind of superiority in the fact your wallet could turn a lump of coal into a Tiffany diamond is annoying, and the fact that you are in a MSF program makes you feel like you have a better understanding of allocating capital than people who have 50+ year track records doing things a "traditional way" is blatant narcissism.

There are lots of ways to make money. I acknowledge that the people at RenTech are brilliant, and they deserve to charge exorbitant fees. Despite this, I would trust my capital to Klarman, Einhorn, or Buffet in a heartbeat over any quant fund. The guys at LTCM were the absolute best, multiple Nobel prize winners, the supposed best risk management in the business, etc. Baupost and Berkshire and may never have years where they are up 50% going forward, but barring a nuclear disaster I'm positive I'll never wake up one morning and see BRK quoted at 0, or that Seth Klarman will have to write a letter to his investors saying that he blew up. Finance at the end of the day boils down into two really simple things, buying a piece of something or loaning money, with an incredible amount of human, industry specific, and macroeconomic factors converging. There are many attributes and risks that can't be modeled. How would you model the fact that Aubrey McLendon is a scumbag who secretly had favorable financing terms to invest in the very best wells after initial output data was collected but not yet published? How would you model that Lehman's disclosures and footnotes were shadier than financial statements of a Camden pawn shop? How does a computer recognize that a business has incredible pricing power because the customer base has a largely intangible attachment?

Do you really think that an engineer with degrees in Math/CS/Physics etc could a better private equity operator than Schwarzman or Kravis? Buying a controlling stake in companies often requires having to deal with executives that are very strong-willed and resistant to change. Is the quant super hero going to soothe the ego of the CEO, negotiate with the labor union, identify which customers to target, which ones to abandon, work with bankers to sell the thing in 5 years, etc. I'll bet every nickel I have that PE will not be dismantled and replaced by a bunch of guys with quant backgrounds. It is an incredibly people intensive business. I'd fucking love to hear some quant guy tell Carl Icahn he could do a better job of turning around a business than him. The guy probably can't even spell heteroskedasticity, but he was up 35% in 2011 and over 20% last year (these are off the top of my head and I was out awhile tonight).

Regarding hedge funds, there are room for quant guys and traditional shops that care about things like quality of management and their incentives, the moat of the business model, pricing power, and a logical explanation for favorable industry economics. If you think that "traditional" hedge funds operate base on the material taught in MBA classes you are sorely mistaken. I've never built a dividend discount model, calculated an F score, utilized Treynor-Black or Black Litterman, modeled an efficient frontier etc in my time at a hedge fund.

I hope you find a job you really like after grad school and make a ton of money figuring out how to use computers to predict the prices of esoteric financial instruments. You are 100%, unequivocally wrong in your quoted assessment though.

Brilliant post!!! +SB!!!

In reply to SirTradesaLot
3/22/13

SirTradesaLot:
IlliniProgrammer:
The large funds tend to be running stat arb and HFT, and yes that requires some programming skill and quant background.

Most of these funds are not stat arb or HFT funds.
http://www.thehedgefundjournal.com/sites/default/f...

Bridgewater, JPAM, GSAM, DE Shaw, Citadel, RennTech all have large stat arb groups. And twenty years ago the rankings were completely dominated by more traditional hedge funds.

I'm not saying they've completely taken over the AUM yet, but you can see the trend.

3/22/13

Do you really think that an engineer with degrees in Math/CS/Physics etc could a better private equity operator than Schwarzman or Kravis?

Do you think electronic execution can really beat a brilliant floor broker?

People believe what they want to believe, and it's tough to believe that the machines are better at making decisions than humans when it comes to finance. We've had a lot of advances in machine learning over the past couple years. It's also a lot more cost effective because machines don't need bonuses or salaries for that matter.

Oh well, I caught more flack two years ago for saying that the trend was probably going to be the same or lower comps and fewer people employed in banking and that there was nothing wrong with saving $20 on a pair of jeans by walking 500 feet. Nobody wanted to believe that comps would continue to remain lower than 2007/2008, but look what happened. Of course, this is a decade-long trend, not a two year trend.

3/22/13

IP please give up. People want to remain idiots and the people who choose to embrace the math/CS while building on everything else will simply crush these fools.

Gray Fox's argument was reducio ad absurdum. The amount of leverage LTCM was operating with, was crazy. It doesn't matter who you are if you play by that style then crazy things can happen.

All the people you listed are old as fuck, I hope you realize. Everyone who is coming into the industry now will be competing with all around businessmen who do know the math and programming on top of understanding EBITDA and P/E. You think you're still in a world where quants hold physics PHD's. You're not, and I can guarantee you that if you hold this hard line mindset of quant vs. businessman you will be the loser.

In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:

People believe what they want to believe, and it's tough to believe that the machines are better at making decisions than humans when it comes to finance. We've had a lot of advances in machine learning over the past couple years. It's also a lot more cost effective because machines don't need bonuses or salaries for that matter.

Machine learning is OK if you have lots of historical data and don't expect too many structural breaks. So, yes, it's great for some forms of arbitrage like statistical arbitrage. But it will not replace human investors anytime soon...

History repeats itself. In the eighties you could hear the praise for expert systems, how they were going to replace doctors, etc. Same hype about neural networks.

It's funny how some engineers are pretty smart but completely disregard the tremendous power of the human brain compared to the machine.

In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:
I'm not saying they've completely taken over the AUM yet, but you can see the trend.

Yet?!? Where do you think this trend is going to stop, exactly?

I'd like to see a market where everyone is doing statistical arbitrage...

In reply to Schumpeter
3/22/13

Schumpeter:

It's funny how some engineers are pretty smart but completely disregard the tremendous power of the human brain compared to the machine.

You have a good point. However, the human brain doesn't scale well and is very expensive.

Many of the far-out predictions made in 1980 about machine learning- facial recognition, driverless cars, research assistance (IBM Watson)- have all come true. Heck, if you have an IPhone, you've met Siri. This stuff was science fiction just fifteen years ago, but it's true today because of better algorithms and better computing infrastructure.

It isn't implausible to have machine learning replace human beings for medium term (two weeks to six months) trading strategies.

3/22/13

And quantum computing will be real in the next 15 years, so I suggest you befriend these computers before they start telling you how to live your life.

And I would argue that most computer engineers/physicists/psychologists in the field of machine learning/cognition/neural networks are some of the men who respect the human mind to the utmost. Think about the real goal of machine learning/AI....It's to model the human thought process. How do you do that? By understanding the human mind. How can you say these people don't appreciate it when that is what they spend their time analyzing and attempting to mimic.

We all recognize the complexity of the brain, I'm under the impression it will never be fully understood under a logical premise. An agent under the logic of a system can not truly understand or appreciate the system since his logic will always become cyclic. This doesn't take away from the fact, however, that a person who knows how to use a computer is by all means superior to a man who does not.

A neanderthal who could use fire was more likely to survive than one who could not.

TLDR;to quote Beyonce:I'm a survivor, I ain't gone give up, I ain't gon' stop up, keep on surviving

In reply to jktecon
3/22/13

jktecon:
IP please give up. People want to remain idiots and the people who choose to embrace the math/CS while building on everything else will simply crush these fools.

Gray Fox's argument was reducio ad absurdum. The amount of leverage LTCM was operating with, was crazy. It doesn't matter who you are if you play by that style then crazy things can happen.

All the people you listed are old as fuck, I hope you realize. Everyone who is coming into the industry now will be competing with all around businessmen who do know the math and programming on top of understanding EBITDA and P/E. You think you're still in a world where quants hold physics PHD's. You're not, and I can guarantee you that if you hold this hard line mindset of quant vs. businessman you will be the loser.

Run some money first then come back and talk. Respect your elders.

In reply to tiger2012
3/22/13

tiger2012:

Run some money first then come back and talk. Respect your elders.


How old are you?
In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:
tiger2012:

Run some money first then come back and talk. Respect your elders.


How old are you?

Old enough to know there is no point in bashing legendary investors. They got there for a reason. This kid doesn't seem to understand the breadth of room for different strategies in the market.

In reply to tiger2012
3/22/13

tiger2012:

Old enough to know there is no point in bashing legendary investors. They got there for a reason. This kid doesn't seem to understand the breadth of room for different strategies in the market.


Since you dodged my question, I'm guessing you're still in college. Weren't you looking for a job at a hedge fund?

I'm old enough to know that strategies and regimes change.

I'm still old enough to remember the CME when there were actual traders on the floor filling out paper tickets. Those people now no longer work as traders at least on the floor of the CME.

Nowadays, most short-term trading decisions in the most liquid markets are made by algorithm. And the holding periods for these algorithms has gotten increasingly longer.

The advantage of age is experience, but the disadvantage of age is that you get set in your ways and you make the assumption that things that worked twenty years ago still work.

It's not going to happen overnight, but eventually, liquid investments will be managed by algorithm. It's a natural consequence of HFT.

I don't think anyone is bashing legendary investors, but it is important to note that their story won't necessarily repeat. In fact, I am arguing that it probably won't. What worked in 1983 won't necessarily work in 2013 and what works today may not work in 2020.

3/22/13

IlliniProgrammer, your initial comment leads me to believe you don't have a very accurate understanding of the breakdown of the hedge fund universe in terms of strategy or AUM. Your cherry-picked selection of major funds includes only one (Renaissance) that's exclusively quant-driven, and at least one that is absolutely NOT materially quant-focused (Fortress). DE Shaw and Citadel have significant non-quant, discretionary fundamental-driven businesses.

There have been many great comebacks throughout history. Jesus was dead but then came back as an all-powerful God-Zombie.

In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:
tiger2012:

Old enough to know there is no point in bashing legendary investors. They got there for a reason. This kid doesn't seem to understand the breadth of room for different strategies in the market.


Since you dodged my question, I'm guessing you're still in college. Weren't you looking for a job at a hedge fund?

I'm old enough to know that strategies and regimes change.

I'm still old enough to remember the CME when there were actual traders on the floor filling out paper tickets. Those people now no longer work as traders at least on the floor of the CME.

Nowadays, most short-term trading decisions in the most liquid markets are made by algorithm. And the holding periods for these algorithms has gotten increasingly longer.

The advantage of age is experience, but the disadvantage of age is that you get set in your ways and you make the assumption that things that worked twenty years ago still work.

It's not going to happen overnight, but eventually, liquid investments will be managed by algorithm. It's a natural consequence of HFT.

I don't think anyone is bashing legendary investors, but it is important to note that their story won't necessarily repeat. In fact, I am arguing that it probably won't. What worked in 1983 won't necessarily work in 2013 and what works today may not work in 2020.

I'm a couple years out of school working as a research associate. Yes I was/am/are/always looking to go buy side depending on it being a good shop/opportunity.

No doubt algos are great for liquidity, short-term periods, and even long-term periods. But there will always be a lot of $ if not majority of $ run on fundamentals. Algos at the end of the day are just extensions of people.

When algos can sniff through an 8-K and put it together that one of the major holders is a forced seller bc they are ceo of another publicly traded company that is getting hammered by einhorn, then i'll be impressed. That should be enough info for a human to figure out what I'm talking about.

3/22/13

IlliniProgrammer, would love to be back in college. Real life's a bitch.

3/22/13

lol at the quant trolls. Nice Friday entertainment.

Just because a company is listed and has a price history doesn't mean an algorithm can determine its value. All the data you can input - financial statements, past prices, news reports, etc. aren't always reliable. There will always be a role for judgement in capital allocation. Arbitrageurs sometimes forget why the securities they're trading exist in the first place, to allocate capital to productive uses. As Schumpeter points out it would be impossible to have a market composed entirely of HFT. Can't happen. Won't happen. Someone has to actually invest.

3/22/13

IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias). The reason I think GrayFox got a bunch of SBs and you got a bunch of MS is that he at least has some respect for the other side of the coin and you have none, even for guys that have been 100X more successful than you could even dream of. There are lots of ways to make money.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

3/22/13

Love the debate and how certain some of the fundamental guys are in their abilities to "analyze" a company.

Shows me the narrow mindedness from both the value investors and quants on this forum

In reply to SirTradesaLot
3/22/13

SirTradesaLot:
IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias). The reason I think GrayFox got a bunch of SBs and you got a bunch of MS is that he at least has some respect for the other side of the coin and you have none, even for guys that have been 100X more successful than you could even dream of. There are lots of ways to make money.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

I want everyone to use the same algorithms/equations, this will stabilize markets. Volatility is always inefficiency, any decent economist knows this.

Gray Fox got a bunch of SB's because there are so many retards who still think basic fundamental analysis is all you need for a future in profiting from these markets.

In reply to jktecon
3/22/13

jktecon:
SirTradesaLot:
IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias). The reason I think GrayFox got a bunch of SBs and you got a bunch of MS is that he at least has some respect for the other side of the coin and you have none, even for guys that have been 100X more successful than you could even dream of. There are lots of ways to make money.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

I want everyone to use the same algorithms/equations, this will stabilize markets. Volatility is always inefficiency, any decent economist knows this.

Gray Fox got a bunch of SB's because there are so many retards who still think basic fundamental analysis is all you need for a future in profiting from these markets.

You watch the close today? Sure that was some serious fundamental volume spike last 90 seconds. 20% of SPY volume flushed. Bet algos had nothing to do with that...

You really want to get rid of volatility? Close the markets, permanently.

In reply to jktecon
3/22/13

jktecon:
SirTradesaLot:
IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias). The reason I think GrayFox got a bunch of SBs and you got a bunch of MS is that he at least has some respect for the other side of the coin and you have none, even for guys that have been 100X more successful than you could even dream of. There are lots of ways to make money.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

I want everyone to use the same algorithms/equations, this will stabilize markets. Volatility is always inefficiency, any decent economist knows this.

Gray Fox got a bunch of SB's because there are so many retards who still think basic fundamental analysis is all you need for a future in profiting from these markets.


Thanks for the laugh. You might want to look up 'Flash Crash'.

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

In reply to SirTradesaLot
3/22/13

SirTradesaLot:
jktecon:
SirTradesaLot:
IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias). The reason I think GrayFox got a bunch of SBs and you got a bunch of MS is that he at least has some respect for the other side of the coin and you have none, even for guys that have been 100X more successful than you could even dream of. There are lots of ways to make money.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

I want everyone to use the same algorithms/equations, this will stabilize markets. Volatility is always inefficiency, any decent economist knows this.

Gray Fox got a bunch of SB's because there are so many retards who still think basic fundamental analysis is all you need for a future in profiting from these markets.


Thanks for the laugh. You might want to look up 'Flash Crash'.

@SirTradesaLot 1. great pun in your name 2. This thread is a lost cause

In reply to tiger2012
3/22/13

Tiger2012:
This thread is a lost cause

Yeah, I give up.

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:
Do you really think that an engineer with degrees in Math/CS/Physics etc could a better private equity operator than Schwarzman or Kravis?

Do you think electronic execution can really beat a brilliant floor broker?

People believe what they want to believe, and it's tough to believe that the machines are better at making decisions than humans when it comes to finance. We've had a lot of advances in machine learning over the past couple years. It's also a lot more cost effective because machines don't need bonuses or salaries for that matter.

Oh well, I caught more flack two years ago for saying that the trend was probably going to be the same or lower comps and fewer people employed in banking and that there was nothing wrong with saving $20 on a pair of jeans by walking 500 feet. Nobody wanted to believe that comps would continue to remain lower than 2007/2008, but look what happened. Of course, this is a decade-long trend, not a two year trend.

Truth.

3/22/13

Difference between us is only one understands both sides.

Do you even remember long division?

It's like a rodent to a man. We both may be in the same world but only one can effectively calculate the others behaviors and actions. I can't explain when you don't speak my language.

Regards

In reply to jktecon
3/22/13

jktecon:

I want everyone to use the same algorithms/equations, this will stabilize markets.

Surely you're joking?!?

3/22/13

You can destroy me in mathematics, right.....I know all the math I need save maybe some more in depth treatments so I could care less.

Oh and econometrics is not math rofl. It is a derivative of statistics... which also is not math, so I somehow doubt your abilities.

The logic systems of this entire universe are cyclical, some are just in lower rungs with faster cycles lol.

And I'm not calling the old guys idiots. I'm saying its time for a more sophisticated investor to dominate matkets just as they did when they came along.

In reply to jktecon
3/22/13

jktecon:

How bullshit is it to hear a finance guy saying a price will go up because the company normally trades at some multiple of its ebitda...That shit is dead. Those people were idiots. I want to hear there is a 65% probability that the price goes up by 10-15% using current macro/industry/company metrics. a computer calculates all this in polynomial time (CS people can correct me if im wrong). And given that probability and our risk tolerance this is the amount that should be put into play to keep risk levels steady.

jktecon:
Difference between us is only one understands both sides.

Do you even remember long division?

It's like a rodent to a man. We both may be in the same world but only one can effectively calculate the others behaviors and actions. I can't explain when you don't speak my language.

Regards


Let's start with your first statement, in which you essentially call the greatest investors of our time a bunch of complete idiots. So you are some punk-ass college kid and think you are entitled to pass judgement on other investors based on what exactly? Do you even have any experience investing / trading (using whatever methods? I mean fuck have you ever even worked?

Leaving the ad hominem aside, your statement is complete bullshit. Statistics cannot forecast these complex processes with sufficient accuracy for you to be able to rely on this probability with any degree of certainty. We cant even get aggregate GDP forecasts with any degree of certainty, these are not beautiful statistical processes that just work. Econometrics is a pseudoscience at best, social and economic processes dont follow physical laws. Mathematics is beautiful and does work, but it is by definition a self-contained system that is internally valid, but cannot necessarily be translated to the real world in all circumstances, as the real world, quite frankly is a mess and our understanding of most economic and social processes is far too limited.

Moving onto your second point, I majored in econometrics and can probably completely destroy you in mathematics. Most guys at my fund have comp sci / physics / engineering backgrounds as well, its not like tradtl hf guys are a bunch of finger painting consultants that oont know maths at all lol.

At Illini, I had some respect for you as a poster prior to this thread but your statement re hedge fund hiring schmoozing MBAs has disgusted me so much as to have lost of all that. No good hedge fund in this world hires schmoozing MBAs. Some hedge funds hire MBAs, most look down on them. And nobody gives a fuck if you can schmooze, its again more of a downside thats discriminated against, we dont need people that talk shit, we need people that can find the truth, not lie. You can be a fucking walrus wearing a dildo, top-hat and monocle that dropped out of high school, if you make money, your hired. Michael Burry is autistic and has made more money than any schmoozing MBA can ever dream of.

There is a place for the schmoozing MBAs, its called Private Equity.

I just love how one back-office guy and one college student together have called the end of investing and discredited all great investors of our time lol.

End of rant.

In reply to tempaccount
3/22/13

tempaccount:
There will always be a role for judgement in capital allocation. Arbitrageurs sometimes forget why the securities they're trading exist in the first place, to allocate capital to productive uses.

Oh of course- that's where the machine learning kicks in. It's not just stat arb- it's ultimately systematizing strategies and ultimately having the computer make the judgment calls, because quite frankly computers tend to have better judgment than human beings- as pilots, as drivers, and no doubt as investment managers.
In reply to SirTradesaLot
3/22/13

SirTradesaLot:
IP -- you realize that if everyone uses algos, they will cease to be effective because they generally interpret data and patterns in a similar manner and are prone to the same types of problems. The good algos are supposed to be taking advantage of human biases...if humans aren't there, they'll have to take advantage of the biases of other computers. Not so hot.

Absolutely. And then people will move further out the holding period curve- until eventually we achieve relative market efficiency. Then they'll try to apply the same algorithms to activist investing and have companies fly by wire rather than have a human CEO.

All of this is going to make the economy more efficient. By the time all of this happens, I will be retired with $500K saved, which will then be earning 25%/year (just like everyone else's money) with less risk than equities today. I'll only need to tap 8% of that $500K to be happy every year, so the rest will probably be spent on National Public Radio or the Foundation for Free Flight.

What's funny is that everyone is prone to supporting whatever it is they're good at, whether it's quant or fundamental (it's human nature to have a self-serving bias).

Oh I completely agree and I fall prey to this. The question is how much is this me actually believing that computers will replace human beings, or how much of this is me just trolling people who believe that they are God's gift to the investing world. (This is often the case at hedge funds.) And the question is just how much everyone else falls prey to their own biases.

Also, private equity (and especially venture capital) are the least likely areas to be 'quantified'. I don't know where you're going with that one.

Sure. But I think that will be the next focus after longer-term investing goes quant. Finally it will be VC. VC is tough because it's the business of figuring out what human beings truly want. PE is easier to systematize. A lot of PE is (1) Find and acquire failing company with a bloated cost structure (2) fire the management (3) cut costs without hurting the customer experience or necessary R&D (4) sell profitable company with reduced costs at a profit.
3/22/13

So, best case scenario, we're going to be cyborgs that live forever and worst case scenario, it's going to be 'Maximum Overdrive'?.
http://en.wikipedia.org/wiki/Technological_singula...
http://www.imdb.com/title/tt0091499/plotsummary?mo...

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

3/22/13

Okay IP, and the other guys here...I want to take this thread in a different direction, as the horse has gotten the shit beat out of it...

OUTSIDE of finance, automation and technology is rapidly replacing the human workforce. First it was blue collar jobs, now it is moving into the white collar realm. IBM Watson is already a better diagnostician of lung cancer than the top docs at MSK Cancer Center. As we move into a post-scarcity society...how will our current economic model survive? Will we have a hybrid system with some kind of guaranteed income or is it going to be like fucking Blade Runner?

I am genuinely interested in predictions/insight...

Please don't quote Patrick Bateman.

In reply to SirTradesaLot
3/22/13

SirTradesaLot:
So, best case scenario, we're going to be cyborgs that live forever and worst case scenario, it's going to be 'Maximum Overdrive'?.
http://en.wikipedia.org/wiki/Technological_singula...
http://www.imdb.com/title/tt0091499/plotsummary?mo...

LOL, I don't know. I'm not a wild-eyed optimist like some people out there claiming this will happen in 25 years, but I think I'll live to see the day when computers become smarter than human beings at everything. Given that I'm in my late 20s and have genes for longevity, we could be talking 50-60 years. I think the days of computers making better business decisions than humans could be 20 years off, but the days of computers making more beautiful artwork or better movie plots could be more like 50 years away.

I think we'll become a leisure society. Where work is no longer necessary and where elitism, if it exists, is merely based on wealth. Given that most countries are democracies, and given that there is less penalty for redistribution when computers are running everything except the distribution mechanisms anyways, elitism might not be tolerated as much.

It will be very interesting to watch. Oh well, if the machines take over and kill me, I won't be able to worry about it.

In reply to DBCooper
3/22/13

DBCooper:
Okay IP, and the other guys here...I want to take this thread in a different direction, as the horse has gotten the shit beat out of it...

OUTSIDE of finance, automation and technology is rapidly replacing the human workforce. First it was blue collar jobs, now it is moving into the white collar realm. IBM Watson is already a better diagnostician of lung cancer than the top docs at MSK Cancer Center. As we move into a post-scarcity society...how will our current economic model survive? Will we have a hybrid system with some kind of guaranteed income or is it going to be like fucking Blade Runner?

I am genuinely interested in predictions/insight...

I think there's a lot of natural stuff in place to ensure redistribution when Gini coefficients get high enough. The economic cost of redistribution gets a lot lower when ginis are high, and the political desire for wealth redistribution also goes up when you see guys blazing by in Ferraris but when it comes to your means of transportation, your health insurance won't even cover crutches.

I'm a Malthusian. In the long run, I think we need to make it difficult to have 37 kids with seven different partners. But otherwise, I think people are too educated and society is too strong for us to have a dystopian solution. (However it's possible, and the reason we won't have a dystopian solution is that people believe dystopia *could* happen.)

Computers are going to make a lot of things cheap. One thing that will be expensive for a long time is going to be ideas and content and creativity.

In reply to IlliniProgrammer
3/22/13

IlliniProgrammer:
DBCooper:
Okay IP, and the other guys here...I want to take this thread in a different direction, as the horse has gotten the shit beat out of it...

OUTSIDE of finance, automation and technology is rapidly replacing the human workforce. First it was blue collar jobs, now it is moving into the white collar realm. IBM Watson is already a better diagnostician of lung cancer than the top docs at MSK Cancer Center. As we move into a post-scarcity society...how will our current economic model survive? Will we have a hybrid system with some kind of guaranteed income or is it going to be like fucking Blade Runner?

I am genuinely interested in predictions/insight...

I think there's a lot of natural stuff in place to ensure redistribution when Gini coefficients get high enough. The economic cost of redistribution gets a lot lower when ginis are high, and the political desire for wealth redistribution also goes up when you see guys blazing by in Ferraris but when it comes to your means of transportation, your health insurance won't even cover crutches.

I'm a Malthusian. In the long run, I think we need to make it difficult to have 37 kids with seven different partners. But otherwise, I think people are too educated and society is too strong for us to have a dystopian solution. (However it's possible, and the reason we won't have a dystopian solution is that people believe dystopia *could* happen.)

Computers are going to make a lot of things cheap. One thing that will be expensive for a long time is going to be ideas and content and creativity.

As someone whose spent more than a few nights at Kam's and pissed on the Alma Mater statue...
As someone whose had great discussion about life in Everitt Lab and Siebel with my friends who got their undergrad AND masters in electrical engineering...
As someone who likes old, shitty, RWD, manual transmission cars (no Civic for me)...

I gotta say...

You say some really strange things.

Follow me on Twitter: https://twitter.com/_KarateBoy_

3/22/13

The guys who keep saying how quant funds take over the world clearly have never worked at a fund manager (of any kind). If you've ever worked on a real quant strategy it should immediately be evident to you why you can't apply quant strategies to everything.

3/22/13

In addition, anyone who has done any work with machine learning algorithms should pretty quickly realize that they have severe limitations.

3/22/13

Ah the old "quant funds are superior" thread returns to WSO...a true classic right next to the "i have a 2.5 gpa from Muhlenberg College, how do I get a banking job at GS?"-thread. This one has been on repeat since 2004.

truth be told...quantitative startegies are mostly just cleverly marketed short vol and they do very well until vol spikes. in 2008 they got destroyed while qualitative macro funds raked in the dough. Some of them are better then others, just like with any strategy, but there is nothing magical going on in the quant world and the returns bear that out. As usual IlliniProgrammer feels free to sput off about stuff he knows less then nothing about, but I can tell you as someone who has worked in the business of running money for over a decade that the same stuff about quant funds being superior to discretionary funds was being said back when i started out and since then many people have gotten rich while the IPs of the world have focused on getting more degrees and saying those who arent Phd level programmers are wasting their time. Thank god i didnt listen to these people.

Learn some basic programming to help you create good spreadsheets but do not become a programmer...the person who makes "the big call" at a hedge fund gets paid alot of money and isnt because they know how to code its becaus ethey have spent years learning how to manage risk and how to trade. The pigeonhold "programmers" are commodities who automate or invetsiaget the ideas that bigger picture thinkers can create....at least at most hedge funds. (obviously ur getco, rentech, etc places are diffierent).

In reply to DBCooper
3/22/13

DBCooper:
Okay IP, and the other guys here...I want to take this thread in a different direction, as the horse has gotten the shit beat out of it...

OUTSIDE of finance, automation and technology is rapidly replacing the human workforce. First it was blue collar jobs, now it is moving into the white collar realm. IBM Watson is already a better diagnostician of lung cancer than the top docs at MSK Cancer Center. As we move into a post-scarcity society...how will our current economic model survive? Will we have a hybrid system with some kind of guaranteed income or is it going to be like fucking Blade Runner?

I am genuinely interested in predictions/insight...

Just FYI every technological advance since the dawn of time has displaced some workers, and every one has created this same debate. In the 1600s in england there was whole movement of workers breaking machines because they thought the industrial revolution would make us worse off for the same reason. In reality it lifted millions out of pverty. Technology and automation may make some profesions obsolote but at the same time it frees up human beings to move on and create new opprtunities...it has been that way since man invented the wheel and didnt have to spend most of their energy dragging shit across the ground....my guess is that back then the caveman draggers worried that the would be out of jobs. Automation and increased productivity are our friends.

In reply to xqtrack
3/22/13

xqtrack:
The guys who keep saying how quant funds take over the world clearly have never worked at a fund manager (of any kind). If you've ever worked on a real quant strategy it should immediately be evident to you why you can't apply quant strategies to everything.

The problem is no one here even knows what quant really even means...You got someone on here telling me they studied econometrics which somehow should imply that they are skilled in mathematics; like wtf are you talking about?

I'm certainly not implying that a straight up programmer with some math is going to take over finance in the next 10 years. I am saying it is a new age in business, where people with skills in programming and mathematics on top of the other business knowledge will be at an advantage.

People who only have knowledge of business seem to think that having an awareness for higher level maths/Cs/econ theory are automatically not fit for finance or mainstream business. This is false, and dying fast. You can either embrace it or get run over as people begin to utilize the new tools available to them.

I really don't see where there is even an argument; like how retarded can you be. What would you think of someone in this day and age attempting to get a job as an analyst without skills in excel or knowledge of a DCF model (which is comprised of algebra). Do you think it is impossible that a better model comes around that perhaps takes some stochastics into account? Or maybe incorporates new theories from probability/analysis?

Math is really symbolic logic at its heart. If you look at many math texts the symbols don't even remain constant mainly just the logical implications. If you can't embrace logic then it necessarily implies that you are illogical. Business is not illogical at the end of the day so if I have these tools, utilize them, and still communicate as well if not better than you, please convey to me why you should be hired.

In reply to jktecon
3/23/13

jktecon:
xqtrack:
The guys who keep saying how quant funds take over the world clearly have never worked at a fund manager (of any kind). If you've ever worked on a real quant strategy it should immediately be evident to you why you can't apply quant strategies to everything.

The problem is no one here even knows what quant really even means...You got someone on here telling me they studied econometrics which somehow should imply that they are skilled in mathematics; like wtf are you talking about?

I'm certainly not implying that a straight up programmer with some math is going to take over finance in the next 10 years. I am saying it is a new age in business, where people with skills in programming and mathematics on top of the other business knowledge will be at an advantage.

People who only have knowledge of business seem to think that having an awareness for higher level maths/Cs/econ theory are automatically not fit for finance or mainstream business. This is false, and dying fast. You can either embrace it or get run over as people begin to utilize the new tools available to them.

I really don't see where there is even an argument; like how retarded can you be. What would you think of someone in this day and age attempting to get a job as an analyst without skills in excel or knowledge of a DCF model (which is comprised of algebra). Do you think it is impossible that a better model comes around that perhaps takes some stochastics into account? Or maybe incorporates new theories from probability/analysis?

Math is really symbolic logic at its heart. If you look at many math texts the symbols don't even remain constant mainly just the logical implications. If you can't embrace logic then it necessarily implies that you are illogical. Business is not illogical at the end of the day so if I have these tools, utilize them, and still communicate as well if not better than you, please convey to me why you should be hired.

it's still not clear why you think that. In general, people on WSO respect the idea of pursuing knowledge but you just sound like a kid without any real world experience.

Why should anyone learn economic? There's a reason why its called a dismal science. I would even argue that trying to predict the economy is detrimental.

What's the value of learning programming? I don't see why having an understanding of how the language of Excel works is going to help me build a better model. You think any CEO/decision maker cares HOW the data was compiled? More importantly, why do you think it would help knowing this fact?

You keep talking about applying higher-level of math skills to incorporate probability analysis. This happens all the time when analysts model to a bear/base/bull case. In the real world, when it comes to finance and company analysis (obviously not engineering) math above simple algebra is more often a safety blanket than anything else. Coming out with probabilities based on standard deviations or stochastic processes causes more problems than it solves.

None of the above is a dismissal of anything you're advocating. But it is a proclamation of its limitations, which you don't seem to be aware of.

Follow me on Twitter: https://twitter.com/_KarateBoy_

3/23/13

are central bankers using this dismal science? Good enough for me.

No, no CEO or manager cares about the efficiency with which business is carried out you're right so let's forget programming.

And you're right, options pricing is no longer based on a stochastic differential equation. And what if this just begins to pervade throughout finance. Don't you wonder why these physics Ph.D's ended up on wall street in the first place. Don't you wonder how a hedge fund filled with astrophysicists with absolutely zero finance training is one of the top performers? Great it may be the only one right now, but maybe they figured something out, no?

What if you learned even an ounce of that and coupled it with finance knowledge, would you not be a more impressive candidate?

In reply to jktecon
3/23/13

jktecon:
like how retarded can you be.

Asks the guy who ends questions with a period.

adapt or die:
What would P.T. Barnum say about you?

MY BLOG

In reply to jktecon
3/23/13

jktecon:
What if you learned even an ounce of that and coupled it with finance knowledge, would you not be a more impressive candidate?
vast fucking difference between that and skynet coming tomorrow to take my job
In reply to SirTradesaLot
3/23/13

SirTradesaLot:
jktecon:
like how retarded can you be.

Asks the guy who ends questions with a period.

Says the troll who doesn't even have the decency as a troll to discern a rhetorical question =X.

3/23/13

The only thing I've learned so far in this thread is that if I ever work for a hedge fund, it'll probably be in investor relations.

In reply to jktecon
3/23/13

jktecon, you are exposing your ignorance here. For a start, the field of econometrics at a high level is really, really, really hardcore. And it includes material from stochastic calculus, probability theory, real analysis, measure theory, functional analysis, information theory, and even machine learning in some cases.

Yeah, it is not high-level abstract math, but in the real world quant funds don't use that stuff; finance doesn't work that way. Pure mathematics has gone far beyond what could be useful in finance in the foreseeable future.

People tend to glorify the math and programming, believing such fields require more basic intelligence to get a degree in. That may be true; I myself majored in a quant scientific field, and I am proud of it. But the more time one spends trying to get accurate predictions from R while studying high-dimensional datasets, or trying to make sense of horrendously noisy macroeconomic time series data with only a handful of years' worth of data, the more one starts to see that one really can only get so much out of statistical analyses. Some problems an algorithm can solve beautifully. Others, not at all, even in apparently similar situations. Even with the best mathematics and coding skills, one can only do so much.

Take go and chess. A decade ago, computers became unbeatable at chess by any human opponent. But in go, computers are still not even competitive at a high level. Maybe one day they will get to the point where they can beat top players at this board game, but right now the top computer programs aren't even close. The games at first seem similar; players take turns making moves from a simple ruleset on a small board. But in fact, the nature of go favors human cognition, whereas chess can be more easily analyzed by a computer.

That said, over the next few years you are definitely going to see more quant tools being used in fundamental analysis. But as someone else pointed out, those tools are going to be made easy enough to use that an Excel monkey can just plug and play.

Also, as Bondarb pointed out, and friends at top quant funds (think DE Shaw, Citadel) have told me, most quant strategies are really just gussied up versions of levering up and shorting volatility, and they always look brilliant right up until they break down completely. Firms that rely on value/momentum approaches have some success at generating alpha, but not that much. Even Renaissance Technologies' "traditional" quant equities fund (RIEF) has had mediocre performance since inception.

And lastly, even now it is not really that hard to get an analyst gig without knowing anything about Excel or DCFs.

3/23/13

Is economics math? It uses almost every field you listed there. There are probably psychologists who are using abstract math as well, and probably political science researchers as well.

I don't believe I ever argued that a straight up quant is going to dominate markets now or even going forward. My argument is that knowledge of mathematics and programming is useful now and will more than likely be a requirement going forward.

Unfortunately I can't speak on any quant strategies so you've struck me dumb there.

I am not too familiar with go but I can say that it is most likely like other games, which should theoretically still be a finite state reality. If so then it is computable, we just need the resources. Quantum computing may strike some of these things down or maybe it will be from the field of automata theory but irrespective of that I argue that a human being should at this point be aware of the technology that they are literally using to argue against me.

Suppose you even do have some great theory (completely free of mathematics/computer science because the fields are "theoretical") and begin a hedge fund or a sell side research firm. What if you are completely unawares of the field of cryptography and are sending sensitive data without up to date security protocols. Could some dumbass programmer in willakootchee Iowa not completely profit off your idea leaving you unawares?

I'm not saying you need to be the network security guy. I am saying you should realize there is a financial risk at hand. And if I'm not mistaken the number one goal of any financier is risk management?

If you logically correct my arguments I guarantee you I'll admit I'm wrong, I am not just going for number one internet troll right now, I embrace knowledge.

In reply to jktecon
3/23/13

jktecon:
are central bankers using this dismal science? Good enough for me.

No, no CEO or manager cares about the efficiency with which business is carried out you're right so let's forget programming.

And you're right, options pricing is no longer based on a stochastic differential equation. And what if this just begins to pervade throughout finance. Don't you wonder why these physics Ph.D's ended up on wall street in the first place. Don't you wonder how a hedge fund filled with astrophysicists with absolutely zero finance training is one of the top performers? Great it may be the only one right now, but maybe they figured something out, no?

What if you learned even an ounce of that and coupled it with finance knowledge, would you not be a more impressive candidate?

This is the last time I'm posting in this thread b/c arguing on the internet has its limits.

From a high level, your arguments are ridiculous and in flux.

You cannot argue that "businessmen" can get an advantage by incorporating "econ/math/etc" and then use the central bankers are your defense of the validity of such a practice. Most people don't have time to pick-up a PhD in econ/math/computer science before applying for a job.

But if you don't think econ is a dismal science than Alan Greenspan found a new friend. The rest of us, including a lot of economist, are going to enjoy the show. What's next, orthodontics will advocate the health benefits of straight teeth?

We were/are talking about futility of applying high-level probability analysis to traditional fundamental analysis and you bring up option pricing? You do realize that a fund can make money off of superior company specific analysis or they can play the structure of the market, through arbitrage strategics and such. right? You do realize that some quant who can more effective in pricing some esoteric instrument or "picks up pennies in front of steam rollers" does nothing to undermine traditional fundamental analysis, right? These guys are fishing in two different ponds and you're comparing apples to hammers.

I reiterate my earlier post. You are acting like a college kid who thinks s/he knows how the world should work. Why would a residential contractor care that you know an OSB panel is manufactured? They won't because that knowledge won't make them money. Same goes for your argument that rudimentary understanding of the foundation of excel would make one a more impressive candidate.

You want to impress me? Show me you understand how the management incentive structure is driving capital allocation decisions.

Follow me on Twitter: https://twitter.com/_KarateBoy_

In reply to jktecon
3/23/13

jktecon:
Is economics math? It uses almost every field you listed there. There are probably psychologists who are using abstract math as well, and probably political science researchers as well.

I don't believe I ever argued that a straight up quant is going to dominate markets now or even going forward. My argument is that knowledge of mathematics and programming is useful now and will more than likely be a requirement going forward.

Unfortunately I can't speak on any quant strategies so you've struck me dumb there.

I am not too familiar with go but I can say that it is most likely like other games, which should theoretically still be a finite state reality. If so then it is computable, we just need the resources. Quantum computing may strike some of these things down or maybe it will be from the field of automata theory but irrespective of that I argue that a human being should at this point be aware of the technology that they are literally using to argue against me.

Suppose you even do have some great theory (completely free of mathematics/computer science because the fields are "theoretical") and begin a hedge fund or a sell side research firm. What if you are completely unawares of the field of cryptography and are sending sensitive data without up to date security protocols. Could some dumbass programmer in willakootchee Iowa not completely profit off your idea leaving you unawares?

I'm not saying you need to be the network security guy. I am saying you should realize there is a financial risk at hand. And if I'm not mistaken the number one goal of any financier is risk management?

If you logically correct my arguments I guarantee you I'll admit I'm wrong, I am not just going for number one internet troll right now, I embrace knowledge.

I shouldn't be continuing this discussion. Your thoughts on the future of quants have been pretty well addressed at this point. But I'll do one more; this will be my last post in this thread.

(Responding to your paragraphs one by one)

1. You're being disingenuous. Training in advanced econometrics requires training in advanced mathematics. Real econometrics is nothing like what you saw in your econ/finance pre-req course. Someone who has a strong background in econometrics also has a strong background in mathematics, which is the relevant point here; the econometrics guy can lecture at you about math because he knows math. In addition, if you look at quant funds' job postings, they often explicitly ask for econometricians. That's the sort of people doing most quant work.

2. Be careful about proclaiming the dawn of a new era in quant strategies before you've spent much time reading about them/learning the basic techniques behind them, at least.

3. Chess, computer scientists agree, could not be solved with a normal computer within the next couple million years. A quantum computer may be able to, but even that is not clear. Go has even more positions than chess, and even a quantum computer would struggle to fully solve it. Eventually, a computer may surpass human players at go, as better heuristics are designed, but it will be a while yet. Maybe decades. And that's the thing: go is simple! it's a small board, with a finite number of moves, and still a computer struggles. Finance is madness, with obscene numbers of variables, as well as human unpredictability. There are no steady trends; the trends break down. Quant trading has had its best success at extremely short time-frames where human decision making is basically irrelevant. Can a computer pick out what business will perform better over the medium term from available data sources? The most successful computer approaches right now rely on buying a large bucket of companies with some grouping of valued characteristics, such as a long-term historically high return on equity or what-have-you. But no quant fund has made any real progress in predicting the future of individual companies, and no statistician who isn't a utopian dreamer thinks such progress is likely.

4. Actually I do know two traders who have long relied on successful strategies that would be replicable if found out by others. They lock their doors, keep a low profile, and don't do sensitive work on computers connected to the internet. Problem solved, in an elegant and cryptography-free way.

3/23/13

To end this thread on a lighter tone, and since we're talking about cryptography:
http://xkcd.com/538/

I think it actually illustrates fairly well the disconnect between nerds and the real world.

3/23/13

jtk just fits talebs definition of nerd so perfectly its almost beautiful.

3/23/13

As long as humans are at all involved in markets then they will be unpredictable.

Hope everyone is enjoying this POMO "just the tip" moment in our centrally planned lives. Selling vol is in vouge again.

3/23/13

Alright. So a non programmer can be a fund manager in the years to come right? Thats all i was asking.

In reply to knowledgeispower
3/23/13

knowledgeispower:
Alright. So a non programmer can be a fund manager in the years to come right? Thats all i was asking.

The answer is a resounding yes.

In reply to Schumpeter
3/23/13
3/23/13

And I am saying that the future of this business has always simply been reliant on knowing what the majority of people are doing. That is all that it is. The black-scholes equation gets brought up because it was not just theoretically right, it was pretty accurate when tested in the market even before current convergence to it, so it showed the possibility is there with other instruments. And to this effect we continue to use such equations in derivatives pricing.

I am arguing that this financial world will almost without doubt be the first space in business to be completely filled with mathematics. Just out of human beings desire to have knowledge that can be tested scientifically. Once there is higher ratio of mathematical based pricing vs. fundemental/technical analysis based I would have to suggest that you jump on board.

I'm not saying everyone should know cryptography or measure-theory based differential equations with Lebesgue measures and the like. But every financier now in my opinion should know math around an engineering level, which really is not too high a level at all. Every human being without a learning impairment (I believe at least) is capable of learning engineering level math and basic programming.

In reply to Schumpeter
3/23/13

Schumpeter:
To end this thread on a lighter tone, and since we're talking about cryptography:
http://xkcd.com/538/

I think it actually illustrates fairly well the disconnect between nerds and the real world.

yeaaaaa...except again you don't even understand that RSA is a cryptographic protocol that relies completely on Abstract Algebra and number theory. No cryptographic protocol is ever proven secure there are just upper bounds found given certain limitations.

And it is known that in this case the limitation of cryptography is that you assume your attacker is not able to solve the discrete logarithm problem. Which may be getting solved quite soon so there's that.

That was my addition to the joke so please laugh at my expense because that's what I think when I see that cartoon.

3/23/13

Just so you guys know this same debate was going on when i graduated in 2002 with the same group of people saying that finance would soon be a battle of robots. It reached its zenith in 2006-2007 when vol was extremely low and tons of quant funds popped up...back then to raise money in a discretionary strategy one literally had to make up some BS that connected it somehow to the quant World. Then in 2008 quant funds got obliterated relative to discretionary funds and many of them including big names like GS Global Alpha were exposed as being highly ordinary and mostly involved in pretty simple strategies. Today i am surprised quant funds arent more in vogue given how low volatility is especially in fixed income, and my guess is that they will continue to grow in popularity until the next time vol spikes. It's just another strategy and when vol is low and a historical data set is a good aproximation of the future then computers do well finding inefficiencies, but when vol spikes and the Worlds changes rapidly those datasets only pick out regime shifts that should not be faded or they find good trades but employ levels of leverage that are inapropriate and get stopped out before the tardes can work.

Pages

What's Your Opinion? Comment below:

Login or register to get credit (collect bananas).
All anonymous comments are unpublished until reviewed. No links or promotional material will be allowed. Most comments are published within 24 hours.
WallStreet Prep Master Financial Modeling