Do It Yourself Masters in Financial Engineering
blog.hiremebecauseimsmart.com
blog.hiremebecauseimsmart.com
Counter argument: Creativity entails not only novelty (that outside the box stuff) but also utility, and in order to be useful, it has to go above-and-beyond what is already known (that inside the box stuff). It's about embracing what's been done before, because it's practice that gives us the capacity to build on the old ideas to make the new.
If you never venture outside the box, you will probably not be creative. But if you never get inside the box, you will certainly be stupid.
If you never get inside the box, you won't even know what the box means or what "outside the box" means.
I think I will use this in the future.
1. Sign up for an Interactive Brokers account; state that you want to trade options.
2. Transfer >25K (day-trading capability) or whatever is reasonable amount money for you. Enough for you to care about but not over the amount where if you lost it all, you'd lose sleep over.
3. Now use the highest/lowest volatility scanner and buy/sell some options randomly in that list; max out your real cash to accumulate positions (do not go on margin).
4. Now figure out what the fuck you just did and how much money you are going to lose. This is the most important step, nothing motivates you to learn security analysis, risk management, Black-Scholes/Bionomial Tree option pricing, post-transaction cost analysis than the threat of losing your money.
5. If you get over step 4, then you probably have a strategy that turn over 7% monthly, or turn over 100% annually. Now automate it as much as you can with IB API.
1. Go to Las Vegas
2. Transfer >25K in chips (high-roller capability), or whatever is reasonable amount of money for you
3. Play the tables
4. Figure out what the fuck you're doing while you're doing it
5. If you get over step 4, you probably think you have a solid strategy for winning against a casino. Now, automate the task by getting people to pay you for your new book/DVD and public talks.
There are plenty of people who should not be in the stock market, and are basically just gambling.
But in the casino table games, in general, the odds are always against you. There IS no winning strategy.
In a market, any market, there IS a winning strategy at any given time, though the strategy will change with time along with the market.
The point of the PP was more "Pay $25k to go to school and learn about trading stocks" or "How to play with $25k on the stock market to force yourself to learn what's up in real life."
Either way you are out $25k, though in the second scenario there is the chance you may actually end up ahead.
Options can't be bought on margin. They lose value over time and thus can't be used as any form of collateral.
(It's not that they try to get you there - it's just a trading platform for pros. A friend once told me how part if portfolio was sold automatically because of an automatic margin call when he didn't even realize hew as on margin)
Of course, to make that work, you've got to convince people with lots of money that you'll beat the market, else they'll never hand over the cash. These people don't care for your high-falootin math, so you should probably leave out the real gory details when you talk to them, but you can't make it too obvious that you're babying them, you need them to feel like they could totally understand all of it if only they weren't too busy to listen; further, if you know anything about your math, you know that the market's not very easy to beat, and that if they knew how tough it was, they'd never give you their money.
In other words, it's bullshit time: you're gonna have to lie through your teeth if you want that money.
Having lost lots of money for other people in the past is the best way to look credible, though in some cases it's okay if you've come out ahead, too. Be careful though, if you've done too well they might start to wonder how much of your own wad you're putting into this thing, and you never want to play with your own money, so it's always better if you don't have to answer the question at all. Never mind that you took your 1% per year off the top of all the money you lost before you lost it (and really don't mention the 10% of each up year!), at least as far as your future investors know, you don't have any Real Money to speak of. :)
People with lots of money and an interest in mathy funds are often familiar with the two base trading strategy types (well, the two base types other than that boring old "buy good companies and hold them like a puss" strategy), momentum and mean reversion - "mean reversion" in particular makes these guys feel really smart when you say it and they understand what you mean, as does "volatility", so say those words a lot, it will help.
As you construct your pitch, you have to make sure to take into account the recent MFE grad, physics major, or computer geek that your mark...err, "future investor"...will hire as a consultant to look into it and see if he can automate the strategy on the cheap based on the details you've given him. The reaction you want from that consultant is less "this is a really good strategy, I should try to do it!" or "this is idiotic, I'm going to expose this" and more "this looks like a joke, but it's strung together plausibly enough that this rich moron's probably going to give these guys a lot of money, I wonder, can I get a job there in exchange for my silence?".
Taking all that into account, here's the type of elevator pitch you want to shoot for:
"Alpha will be created by constructing synthetic Asian lookback options on emerging market debt with valuations derived from our proprietary distributed high frequency mean reversion algorithm that predicts, with 98% accuracy, volatility skews at the millisecond time scale and below."[1]
Blammo.
[1] Your actual trading strategy, of course, will be "Put as much money on the line as possible without getting sent to jail. Cross fingers, pray for profit after a year. Collect huge pile of money if so, change name of fund and create new, better fund with bigger investors next year if not. Repeat."
Pretty different than trading against an existing market.
Books are free, if you have the internet or live near a university library.
I do like your empirical attitude though.
Here's my reading list: http://personalmba.com/best-business-books/
Here's the book I wrote, condensing what I learned in the process: http://www.amazon.com/gp/product/1591843529/
I now own my own business, work for myself as a business teacher and consultant, have an enormous amount of flexibility in my day-to-day schedule, and make more than professors who teach at top business schools.
Education is changing.
It's not uncommon for me to go to the library or B&N and read a stack of 20-30 books in a few hours. I've been reading business books for six years, and easily read 200-400 books a year.
Here's a post I wrote on effective non-fiction reading techniques, for more info on the approach: http://personalmba.com/10-days-to-faster-reading/
-- 10-20 minutes per book? -- 20 to 30 books in a few hours.
I am a bit skeptical, but maybe it is doable. I am curios to know if anyone else on HN can read/comprehend a non-fiction business book in less than 30 mins.
I went ahead and ordered '10 days to faster reading'.
I could see things getting to the point to where you could have the option to generate an executive summary from any popular book that tries to hit the main points. Then you drill down as you want more information. Books could function almost like mini-wikis at that point but with a narrative or order that you use to drive through it.
2. buy a copy of Neuromancer
3. search for the part where Case tells his Ono-Sendai to "sift this shit. 2 minute precis"
4. Be amazed once again that a guy with a typewriter predicted what you're doing.
Serendipity! I was thinking the same thing -- though I was thinking more along the lines of Quora posts.
This is a feature I have posted about and asked for on Reddit actually; the ability to tag information to topic and quality etc.
This would really apply to quora best though, as the information is already sorted in one dimenstion: quora topics.
I'd like to be able to assemble and curate information from sources like quora and get them into a more of a how-to; infact I posted a question along these lines on quora, with little response.
(http://www.quora.com/Would-Quora-users-like-an-entrepreneurs...)
However - I predict that this will be a golden goose of social search. A question is asked within a particular topic, experts respond - then the best answers to the most popular questions within a topic, and tagged by the audience, are assembled into cliff-notes of information on that topic.
Fortunately, this is what nearly all "business" books are. They repeat that two pages of content over and over. If you spend 20 minutes skimming, chances are you will hit the actual content at least once.
Another interesting point is that most business books, if they cite references, usually only cite a few. Often times one can look up the paper(s) it references and get the actual content in a much more, ah, compact form, without most of the sales.
Of course, /actually reading/ business books rather than skimming them, I think, is a reasonable way to figure out how to talk to and even to sell to business people. I mean, that's what the book is doing, right? even if the actual "content" of the book is complete BS, if the book is a popular one, it's obviously good at selling to business people, and you can learn by example.
http://artofmanliness.com/2009/10/18/how-to-speed-read-like-...
Maybe business books are just different than technical/science oriented ones.
10-20 minutes won't cut it for deep technical material, particularly if the book is designed to teach via working through examples. If you're reading a workbook, then take the time to work through the examples. Choose your workbooks carefully, since they require a more significant time investment.
Most non-fiction books, however, are written around a few simple, straightforward concepts - you can identify them in a few minutes. A few more minutes identifying examples, relationships to other concepts, and counter-examples / edge cases, and you've extracted 90% of the value of the text.
Almost every non-fiction book fits this model. It pays to tailor your reading approach to what you're reading.
As to value - in general, most business books are marginally valuable in isolation. Once you read many of them, however, very useful patterns begin to emerge. Recognizing and using those patterns provides the value.
I'm at the point where I can sometimes skim through arXiv papers and understand the gist.
(Does it work the same for books with code in them? Legit coders can skim and understand?)
Since they're not actually written in PHP, I've got to switch to a new language. ObjectiveC for iPhone, Android-Java, or any number of cross platform languages to pick up a third party framework.
On top of that - there's the issues of structure and stuff.
For learning a new language, there's no way to skim.
HOWEVER for something like an A* pathfinding algorythm (an artificial intelligence method of finding the most efficient path from A to Anywhere) - you can get the "Gist" by skimming a paper. I understand what A* is and how it does it ... but if I want to replicate it I still have to go over it line-by-line if I want to get it right though.
So, I think the short answer is "NO" coders can't skim stuff and get the job done right.
Well, fortunately in that the signal-to-noise ratio is high. Unfortunately in that one must be very selective about what one chooses to invest time into reading.
One quibble: I'd put Steve Blank's "Four Steps" under Entrepreneurship rather than "Value Creation and Design". I've read Four Steps, Universal Principals of Design, and Design of Everyday things (I have not read Rework, the other book in the category). Four Steps really covers much different material than the other other two books.
That said, I'd recommend getting your hands on a hard copy if you can - the book is designed to be equal parts primer and reference, and the hard copy is optimized for easy reference use.
To my mind part of being an entrepreneur is leaving the world of books, professors, and ideas.
Anyway congratulations and here's to hoping that your book makes the B-school reading lists.
you don't pay $100k for the knowledge, you pay $100k for a ticket to the party. everyone knows that.
of course, if you want to be an entrepreneur or trade your own $, that is a different story. but this seems unrealistic for any kind of mainstream career move.
It's actually the reason I started writing on the internet (note the extremely clever site name).
1. Signalling to your future employer / Getting past the HR filter. 2. Networking with others.
Learning something useful is usually a bonus.
My understanding is that your value to a firm is directly proportional to your ability to generate alpha. Credentials would be useful for securing an entry-level position, but eventually, it is your ability to develop profitable strategies that will secure your position in this industry.
This is such a secretive business, that I doubt academia will be teaching anything other than the fundamentals/theory . As pointed out in the article, MFE employment stats are murky at best. I haven't had much luck getting confident stats. There's only a few of programs that staff actual practitioners, people who, you know, actually made/make money doing this stuff.
One of my concerns with the credential-path is future career growth. Many of the firms that hire from schools place candidates in specific roles that need to be filled. You might get stuck in an area thats not as profitable as other areas of the market. You will also not be exposed to the firm's intellectual property until they've worked you hard and for many years.
Contrast this with the option to start independently. I've spoken with quite a few people who've started out this way. Some have parlayed their experience and performance into jobs/partnerships at the big firms. Others continued to scale and build out their own fund. They all say the same thing, if you can develop profitable strategies, you will find people more than willing to invest in your strategies.
As far as networking, I think the internet is helping that immensely. Quite a few fundamental investors have received job offers or started their own hedge funds from blogging/posting on forums when it became apparent they had an edge. There's always a few headhunters and fund managers hanging around nuclearphynance, and other forums.
There is a higher risk, perhaps, of getting nowhere. But you also have the freedom, and less dogmatic structure to think independently, and the time to learn things thoroughly (vs. in a degree program).
I'd like to hear other's thoughts on this.
But just like in other businesses, perceptions, measurability, and org structure are very important.
It's not as simple as "can you generate alpha". Does your desk manager give you as much leeway as you desire? Does your alpha come only after your position has been underwater for weeks and you were given position to double down?
Measurement is also a huge issue. How much did you contribute to the bottom line? Can you prove it? What if you told Moira about low-discrepancy sequences, which improved her Monte Carlo sims, which somehow affected someone's trading down the line ....?
I met my current (quant fin) employer on the internet. While I was living in a trailer park.
Ironically, governments throughout the world still struggle to find solutions to education. Please, just teach people to open a book and do some damage.
On the other hand, if you're conditioned to believe and trust in the transmission of knowledge through independent reading, like most autodidacts are, you perceive them as an opportunity to improve.
I, myself, am an autodidact programmer. I had never heard of many resources before encountering them in threads on HN, Reddit, or Stackoverflow. I haven't read all books, or followed all the videos, but at least, now that I'm aware of them, it prompted me to start working toward shifting a few things in my life to allow me the time and opportunity to dig deeper in the material.
It's a process.
Real reading happens for me when I shut the computer and focus on just one thing.
The diffusion equation. It was so underwhelming that I lost any interest in going into financial engineering.
If you think that makes you smart, then I want you to be my competitor.
Reddit has a lot of such things.
1. Just because the material is freely available on the internet does not mean you will have the ability to learn from that material. However, an inability to learn the material does not necessarily indicate a disinterest or an ineptitude. Quite frankly, it took the threat of a grade and the commitment of paying tuition for me to sit down and read through Shreve's Stochastic Calculus books, despite the fact I had picked them up almost a year before attending CMU.
2. Reading a bunch of books doesn't make you knowledgable on a subject. Deep interaction with the material does. You don't go to school just for a piece of paper -- you go for a structured learning environment. I went because I believed, and still do, that the teachers who understand this material will know the best way to organize and present it to me. Furthermore, most MFE programs stress the importance of an internship, giving students pragmatic experience. Finally, most MFE programs also offer speaker series to provide interaction with industry practitioners.
3. Whoever is saying 'Girsanov's Theorem' doesn't understand the connection. Girsanov is only relevant from the Fundamental Theorems of Asset Pricing, which is only relevant in that it translates the question of arbitrage into a question of existence of a risk neutral measure. Ultimately, pricing is merely replication.
4. "So you’re trying to get a job where you face the market everyday and disagree with it." This quote tells me you fundamentally are misunderstanding what occurs on most Sales & Trading desks, which is where most people from MFE programs try to place. Don't worry -- most people don't understand what they do. Quite simply, most Wall Street traders do nothing other than execute trades for clients and try to match and offset the risk of the trade with other clients. Sometimes traders will carry some risk in their book (most of the time, it is second-order risk) -- but for the most part, they aren't actively taking bets on market direction (though, rumor has it Goldman is notorious for this…) -- they leave that up to the hedge funds.
5. This whole notion of independent thought versus 'indoctrination' is ridiculous. I'm sorry, but math is math, whether you read it on your own or are taught by a professor. Independent thought is a characteristic of the student, and they either have it or they don't.
6. Is the degree expensive? Heck yes. Is it worth it? Only you can answer that. For me, having the degree is priceless because I own my business, and it acts as a certification of qualification when I meet with clients. It serves as one more point for them to check off in their due diligence on me.
7. You will completely miss out on the opportunity to network with peers. I now know people who I worked with on a daily basis at almost every major Wall Street firm, from investment banks to hedge funds to consultancy firms. That is a priceless network.
So can you learn the material all on your own? Sure -- but I bet it takes you a lot longer than the year and a half it took me. By going to school, I was able to dedicate myself full time (homework and classes well exceeded 40+ hours a week), to learning the material. In my opinion, this isn't really material you can learn 'part time,' and if you try, you certainly won't have the opportunity to develop any practical understanding of it.
I addressed #1 in the article. Suffice to say I disagree.
#2 Who says you interact deeply with material in school?
#3 comes from Emanuel Derman's interviews of fresh grads.
#4 I do understand the difference but I would rather prop trade than make markets.
#5 There are so many sub-areas. Should you study wavelets? Finite element method? Differential geometry? EDA? Bayes? ML?
#6 Perhaps I am more risk-averse than you. Or less trusting.
#2, #7 - I said that's the irreproducible part of an MFE.
Would you mind adding your comments to the Disqus on the site? They're very thorough and as you're an MFE grad they counterbalance what I said. I think your viewpoint is quite valuable.
It was just something "for your consideration". Nothing accusatory.
BTW, when I wrote this piece it was intended for QuantNet.com. However I pissed off the guy who runs it and now he thinks I'm a douchebag.
That's all to say, I never thought this piece would get any traffic at all. The fact that so many people are interested makes me even more skeptical of the putative value of an MFE....
I've been mulling over a project to deliver more courses in a similar vein, but I'm still struggling with how I'll manage to collate the user-submitted YouTube-videos automatically without mandatory approval on my side.