What are some good books/papers to read for someone who knows little about quantitative finance but would like to know more. More specifically, what are good books for those who have a solid math and cs background.
You need to be a little more specific regarding your interests and background. If you are interested in theory, there are plenty of books discussing specific models. However, most practical or buzzword (e.g. HFT) books are rubbish (think about it: if you have a profitable operation, why would you write a book? Unless, like me, you have a screw loose :P )
If you want some practical discussions, "Trading and Exchanges" by Larry Harris is a bit dated, but nothing comes close. Most of the descriptions are still valid today (even if some of the mechanics have changed)
If you want something more along the lines of what a financial engineer would know:
- "Options, Futures, and other Derivatives" by John Hull (goes over basic models without delving too deeply into theoretical math aspects)
- "Stochastic Calculus for Finance II" by Steve Shreve (goes over the basic stuff but has enough stats to keep a grad student happy)
- "Monte Carlo Methods in Financial Engineering" by Paul Glasserman (much more practical, and goes over subtleties of monte carlo simulation and other stuff like low discrepancy sampling)
- "Modelling Fixed Income Securities and Interest Rate Options" by Robert Jarrow (walks through how to perform certain simulations, covers lots of little details most theoretical books skip)
If you want something more theoretical:
- "Introduction to Stochastic Calculus Applied to Finance" by Lamberton and Lapeyre (Nice little intro)
- "Arbitrage Theory in Continuous Time" by Tomas Bjork (Slower discussion, larger breadth)
- "Brownian Motion and Stochastic Calculus" and "Methods of Mathematical Finance" by Karatzas and Shreve (Solid theoretical foundation, for the more mathematically inclined)
There was a good non-measure theoretic discussion of financial models but the name escapes me ATM.
"Analysis and Use of Financial Statements" by White, Sondhi and Fried.
It has nothing to do with computational finance but knowing more than nothing about financial statements and how to think about what a company actually IS can only help you in the long run. Also, it's quite interesting.
It really depends on which angle you're coming from, when you mean "hacker".
Quantitative finance is all math, so in order to really understand it, you need a solid background in probability and statistics. Even the most basic concepts make heavy use of probability, so without strong fundamentals you won't get very far.
There are other concepts, such as risk-free returns, etc that also play a big part in quant finance. A book that I really enjoyed that gave a good background in a lot of these concepts was "Trading Strategies for Capital Markets".
One of the basic concepts of quantitative finance revolves around the Black-Scholes equation, which calculates a price for an option. I would suggest first looking this up as well as how it was derived, to see if you want to pursue this further. If you're having a hard time with this, then quant finance may not be what you're looking for.
A video tutorial website that seems decent for quant finance is Nathan's Lessons:
If what you REALLY mean is you want to program trading algorithms, that's a bit of a different beast. For that you really need to understand how algorithmic trading works, how the market microstructure works, how to place trades, etc. For this there are a bunch of books but the best introductory book is definitely "Trading and Exchanges: Market Microstructure for Practitioners". It's a bit old, but still largely relevant and gives you a lot of the history. I really wish they would update this book, because I would buy it again. Another decent book is "Algorithmic Trading and DMA".
I am a quant. I work at a bank. We do a fair amount of portfolio optimization & price-analytics, in Scala, Matlab,C++, CPLEX. Its not hard to get your feet wet with financial modeling, though reasoning about the models does take an MFE or comparable education.
For example, here's a tiny scala pricer I coded up in 5 minutes to spit out the price of a google at-the-money call expiring 47 days from now, on Dec 17 using a 10k Monte-Carlo simulation.
val risk = 0.28 // google has a 28% implied vol
val T = 47.0/365 // 47 days annualized
println((1 to 10000).toList.map(_=>Random.nextGaussian).map(x=>595 * exp((-risk * risk/2) * T + risk * sqrt(T) * x)).map(x=>max(0,(x-595))).sum/10000)
Output: $23 & change ( its selling at about $23 right now )
To veyron's excellent list, I'd add Brandimarte's "Numerical Methods in Finance".
I'm finding this a more interesting and detailed read than _Capital Markets for Quantitative Professionals_. For instance, the book traces through everything that happens when placing, executing, and settling typical trades.
On the other hand, I've taken the Canadian Securities Course, so perhaps Capital Markets is a better introduction for total beginners.
Are you really interested in learning quant finance (learning how to mathematically model various parts of financial markets, pricing/measuring risk of specific instruments, etc.) or are you interested in programming systems for the financial industry (trading/risk/compliance/banking/etc)?
edit: I just remembered, Paul Wilmott and Hull have several introductory books if you are interested in what is usually called "quantitative finance." This means how to price options, futures, etc. From what I recall, this does NOT mean using statistical correlations to trade two similar stocks, making market-making models, etc.
Exactly. Where I used to work, anyone without a quant background would be urged to buy a copy of Grinold and Kahn the bible you suggested above or "Quantitative Equity Portfolio Management" by Chincarini and Kim.
Any of Paul Wilmott's books. Either his single Introduction to Quantitative Finance, or his three volume set. His focus is on knowing where models work and where they break down, and how to avoid using them naively. http://wilmott.com/
Mark Joshi's (markjoshi.com) site has a pretty good set of information for aspiring quants (in particular, see the "advice for aspiring quants" bit).
The "careers" forum on nuclearphynance.com is also pretty good, and might also shatter some preconceptions about how easy it is to waltz into the industry. (note: nuclearphynance seems to be down at the time of writing)
I'm sure the idea that there are no jobs out there is exaggerated. A MFE graduate can still find a position somewhere in a bank doing something vaguely related to finance. But the days when anyone vaguely familiar with PDEs could waltz into a bank and quickly start earning $300K may be drawing to a close.
There will, however, always be lucrative jobs for those who are truly brilliant and willing to work on stuff which is kinda boring.
The NP posts don't say there are no jobs -- just that junior quant has 1000+ applicants per position.
Jane Street still jingles its bells up at MIT, but for those who are less than six stdev's IQ or don't live in NYC / London, I would say focus on petroengineering instead of quant finance. (Petroengineering is the highest paying college major.)
I saw a bunch of doggie positions at Fidelity for "lower" MFE holders. You can also check up on at least one: Joy Pathak wrote for QuantNet about his experiences doing the Baruch MFE. Jim Liew is running the Alpha Quant Club along with that pack ... those are just a couple people you can stalk.
But I suspect it's just like those who finish a law degree and then can't get a (good enough) job in law: not the university's problem, and they're ashamed they ended up in the bottom 50% of their class.
Ernie Chan's book, "Quantitative Trading: How to Build Your Own Algorithmic Trading Business ", is a good starting point. It's a nice blend of theory and pragmatic practice. Chan provides example implementations in Matlab to help readers get started.
The best resource for mastering probability, and financial mathematics are Yufeng Guo's guides to passing actuarial exams. He takes the hacker approach to teaching fairly difficult concepts so you can focus on solving problems quickly. Most of the books recommended here are too traditional in their approach,(boring) this is a true hacker guide to learning how to frame problems correctly and then the math needed to solve them. If you are a decent programmer you can figure out the application afterward. I used them to pass the introductory actuarial exams and they are awesome.
Hull is always the first recommendation. After that, it depends what area you're interested in. You could move on to Shreve for more hard core quantity material, or go to Tsay (Analysis of Financial Time Series) for a time series primer. You could try Natenberg for options material or Grinld and Khan for portfolio management. Not too much useful material on HFT out there. Agreed that Harris is useful but out of date.
High frequency trading and quantitative trading aren't the same thing. High frequency is very basic math and a lot of technology to make trades happen as fast as possible. Low latency becomes difficult when you're running complex math on real time data. HFT is generally used by market makers. HFT is the next evolution of day-trading, since all trading is trying to react to news quicker.
Quantitative finance is very complex math used to model and backtest data. It is used by hedge funds, long-term traders, and shops that analyze data to find securities that they believe are mispriced.
(This is obviously a highly simplified description of the field.)
Disclaimer: I'm not a quant and don't know where to source data on this topic.
But from what I'e read, the size of the derivatives market was approximately 5-6X global GDP in the mid 00's. Now, it's over 20X global GDP. If that market ever crashed, we'd be looking back at the crash of '08 as 'good times'.
those numbers are really bad measures of risk. those are probably notionals on derivatives. if you're trading a swap, the notional may be 100B, but the amount actually at risk may be something like 10 bps, which is just 100MM.
Also, its not like if the market crashed, all of the value of the derivatives go to zero. Derivative holdings always have counterparties. If you make 100MM, someone else loses 100MM and vice versa. this isn't foolproof, as there is counterparty risk, but it makes the system a lot more stable than your statistics make it seem.
TLDR: The total risk of the derivative market is nowhere near 20X GDP.
I'm not challenging you on this, but I still remember Buffet's comment from years ago that they were weapons of mass destruction. What is the risk and how do regulators ensure that any untoward risk is mitigated?
Buffett makes a great deal of cash flow from writing options, so take what he says with a grain of salt.
"The brilliance behind Buffett’s investment in Berkshire is astounding. He effectively used (and uses) Berkshire as the world’s largest option writing house. The premiums and cash flow from his insurance business created dividends that he could invest in other businesses. But Buffett wasn’t just buying Coca-Cola and Geico as many have been led to believe. Buffett was placing some (short-term AND long-term) complex bets in derivatives markets, options markets, and bond markets."
Thats kind of misleading. First those figures are probably notional underlying, which doesnt mean much. Second contracts have a buyer and a seller, so net value is zero. So it does not work like a stock market crash destroying (presumed) wealth.
They should be teaching kids the basics of this stuff in schools...
Don't forget about counterparty risk. This is something everyone forgot about during 2008 and ended up being one of the biggest problems.
This is what happened with AIG, where people thought they were adequately hedged on their position because they bought a CDS with AIG, and then it turned out that AIG itself had written so many CDSs that it couldn't pay back its obligations, which exacerbated the entire financial crisis.
The news blurbs I see make me think more and more exchanges around the world actually welcome HFT volume. Exchanges in Latin America, for example, are quickly modernizing to allow automated traders to come in.
Remember, exchanges benefit from greater volume (more fees).
With regards to the quants being trashed, you will always need people who can rigorously describe a system (I do realize that rigor and accuracy are different). The quants are not going anywhere :)
I think in general it is governments and regulators who are (attempting/pretending to be) cracking down, not exchanges. Exchanges continue to get faster, and offer premium services to those who need it.