Hot answers tagged

25

I can only talk about quantitative trading. As a rule of thumb, the lower frequency you work in, the more econometrics is important, whereas for a higher frequency, the more econometrics becomes useless. (I would still recommend a top econometrician for HFT since they have what it takes to succeed, it's just the models aren't out-of-the-box applicable.) But ...


12

You may want to first broadly categorize volatility models before comparing between them within each class, it does not make sense to compare standard deviation models with an implied vol model. I would broadly classify as follows: Historical realized volatility: Those include standard deviation (sum of squared deviations), realized range volatility ...


10

I think there are a lot of different ways to specify this problem. For simplicity, consider independent Garch processes $$ r_{1,t} \sim N\left(0,\sigma_{1,t}^{2}\right) $$ $$ \sigma_{1,t}^{2} = \beta_{1,1}+\beta_{1,2}\varepsilon_{1,t-1}^{2}+\beta_{1,3}\sigma_{1,t-1}^{2} $$ and $$ r_{2,t} \sim N\left(0,\sigma_{2,t}^{2}\right) $$ $$ \sigma_{2,t}^{2} = ...


9

GARCH will work if volume has memory with some decay. AR will work if volume has mean reversion properties. Both of these are empirical questions and depend on the market. You should also consider if there are seasonal (day-of-week, monthly, quarterly effects) in which case you would want to add dummy variables. MA models will work well if volume behaves ...


9

Treat the estimate of standard deviation as a random variable. Then you can bootstap the sample estimate and generate t-statistics and associated confidence intervals for your statistics. I describe a generic boostrap process on this post.


9

The best paper is probably Relative Volume as a Doubly Stochastic Binomial Point Process - James Mcculloch. In this paper the volume is modelled via a Point Process, and theoretical laws are derived (with confident intervals, etc). And we put elements about this in Market Microstructure in Practice, Chap 2.1. Volume curves are analyzed, not only during the ...


7

@user2763361 has a very thorough list of useful econometric topics for quantitative finance. I would add missing, mixed frequency, and irregular data as major issues that I'm either constantly dealing with or begrudgingly ignoring. Seasonal adjustment is important too for some data (like electricity futures), though the subject is also related to his ...


6

I basically agree with @John, let me expand: We want to model $y$ using a simple linear model, the most basic setup is $$ y = c + \mathbf{X}\beta $$ with $y$ the $N$ observations, $c$ a constant, $\mathbf{X}$ the $N \times M$ matrix of regressors and $\beta$ a $M$-dimensional vector of coefficients. This model has $M$ parameters, the elements of $\beta$. ...


5

Actually you should be interested by the Berry Essen's theorem which precises the rate of convergence of the central limit theorem. Given i.i.d. $X_1,\dots, X_n \sim X$ 1) GLN : assuming $E(X)<\infty$ then $\overline{X}_n-E(X)\to 0 $ 2) CLT ("rate" of the GLN) : assuming $E(X^2)<\infty$ then $\frac{\sqrt{n}}{\sigma^2} ...


5

From an academic viewpoint you do not have a lot of choices: The Rosenbaum-Robert approach, the price model with uncertainty zones is a model of trades and duration between trades (implicitly). It is worthwhile to try it. You can also use an Hawkes process, it will have the nice effect of capturing clustering effects on trades. if you want to use ...


5

An AR(1), once the time series and lags are aligned and everything is set-up, is in fact a standard regression problem. Let's look, for simplicity sake, at a "standard" regression problem. I will try to draw some conclusions from there. Let's say we want to run a linear regression where we want to approximate $y$ with $$h_(x) = \sum_0^n \theta_i x_i = ...


5

The idea of skipping a month was already in Jegadeesh and Titman 1993. The key academic paper in this area. Jegadeesh himself (without Titman) discovered a 1-month return REVERSAL effect in 1990, so it makes sense that he would take out 1 month in calculating returns in his later (1993) study. He already knew what happens to stocks that are up a lot ...


4

They are not mutually exclusive. For example, the class you refer to as "econometric" are simply linear regression models that include as factors prior returns or residuals of the return series sometimes with weightings on the observations. You could easily design a neural network with no hidden layers and the same inputs. So each of the econometric models ...


4

Working on trigonometric polynomial decomposition, the first step is to take a big look at Fourier transformation. It is very powerfull, well documented and probably well implemented on your favorite language. It will give you the decomposition of your time series. You can remove highest frequencies, which correspond to noise, to have a good estimation.


4

There is a lot of ways to understand why stationarity allows to apply usual time series analysis. Here is one more. Very often, the theoretical justification of what you do in time series need to be able to identify the mean formula and the expectation: $$\frac{1}{N}\sum_{n=1}^N X_n \underset{N\rightarrow +\infty}{\longrightarrow} \mathbb{E} X, $$ where the ...


4

The return equation is just an econometric equation that models stock returns (or other asset returns) as a function of: (i) intercept (i.e. the average return), (ii) some independent variables/features, (iii) noise that has zero mean and time-varying variance. There are sometimes other things in the return equation too that form more advanced models. The ...


3

Building upon +Imorin answer, you should have a look specifically at discrete cosine transforms. It's a standard approach when trying to express finite sequences as a sum of cosines. I would start from there, especially as it's implemented in every common language (R, Matlab, Python for starters). Only then evaluate if you need more.


3

You could read it like this: The typical change in equity value is equal to the typical change in asset value, adjusted for the probability of the assets surviving. Note that the formula is not specific to Merton models, it's also true for regular options and their underlyings. It's just that volatility of option prices isn't typically a concern in ...


3

Well, the main intuition of the Merton model is that a company's equity can be treated as a call option on its assets, thus allowing for the application of Black-Scholes option pricing methods. Let's consider a company that has assets $A_{t}$ financed by equity $E_{t}$ and a zero-coupon debt $B_{t}$ with face value K, and maturity T. At time of maturity T, ...


3

What you are talking about is called regression using fractional polynomials and it has its merits. The canonical reference is this one: Regression Using Fractional Polynomials of Continuous Covariates: Parsimonious Parametric Modelling by Royston and Altman (1994) From the abstract: The relationship between a response variable and one or more ...


3

I deal recently with some analysis of the Volume time series, daily volume in € for European stocks. I found out that an ARIMA model works well. But, some EWMA could also provide good forecast if it's well parameterized. You can also face some seasonality effect due to macroeconomic events, some you may need to clean you data and treat these days in a ...


3

There are tons of quant related blogs out there, some of which contain relatively sophisticated content, others less so. Have a look at the following, which aggregates blogs: MoneyScience Otherwise I could point you to bank/sell-side research. Have a look at the freely available Reuters Messenger (RM), they maintain channels where you can be permissioned ...


3

From my point of view, dynamic models like the one developped in Relative Volume as a Doubly Stochastic Binomial Point Process - James Mcculloch to provide a dynamic forecast of the volume does not improve significantly the forecasting comparing to a static volume curve forecast using historical data (last month intraday data, and an EWMA algorithm). I've ...


3

2) Alternative to Fama-MacBeth is Fama-French approach. Explanation of difference see, for example, here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1271935 Fama-French approach was used by Carhart (introduced momentum), Pastor-Stambaugh (introduced liquidity), Fama-French themselves (used it to build 5-factor model), and many other (elsevier or ...


3

As an overview, Expected Returns, by Antti Ilmanen, was recommended to me. He has a preference for data over theory, so it will appeal to quants. The book is longish, and got a bit heavy at times, but he covers all the investment products and all styles of investing. The biggest problem might be that it is now 3 years old, and was heavily influenced by ...


3

Volatility changes over time. Even if daily returns are normal, assuming the conditional volatility each day is known, the unconditional distribution of daily returns will have excess kurtosis. For example, if daily returns have a standard deviation of 1%, 90% of the time, and a standard deviation of 3%, 10% of the time, the presence of the high-volatility ...


3

There are a couple of issues with your example. First, for this ticker, there is a problem with the Yahoo price data for the period 2014-11-26 through 2014-12-03 in which the prices drop about 80% and then return to their trend line. This appears to be related to a stock split which Yahoo isn't handling properly and isn't real. Its causing part of your ...


3

Yes, it exists and it is called ccgarch package. You can install that by simply running in R install.packages("ccgarch") and learn more about that on the CRAN relative paper. Moreover, I suggest you to read this lecture hold by the author during an R conference. Hope this help.


2

Try the following : perform the logarithmic transformation of the volume data. check if the transformed data fits the normal distribution nicely. if you are working with intraday volume, then adjust for the seasonality for time of the day effect, if using daily data, in some cases some special seasonalities like expiry day, etc might be applied but it may ...


2

yahoo provides adjusted_close. You could use this to detect splits adjustment_factor = adjusted_close/close change in adjustment_factor = adjustment_factor (yesterday's)/adjustment_factor(today) if this number is less than 0.9 or greater than 1, you have a split



Only top voted, non community-wiki answers of a minimum length are eligible