Tag Info

19

Because of: The (extreme) dominance of noise over signal The prevalence of non-repeating patterns (many of which we know are not going to repeat) A pathetic sample size for cross-validation Regime changes due to exogenous events. These are typically in the cross-val window which makes it even worse. (GFC, financial integration, trade law changes, interest ...

7

I recommend reading Cao, Hansch, and Wang (2004) "The Informational Content of an Open Limit Order Book". They present a simple model for an order-book price called the weighted price ($\mbox{WP}$): $$\mbox{WP}^{n_1 - n_2} = \frac{\sum_{j=n_1}^{n_2} (Q_j^d P_j^d + Q_j^s P_j^s)}{(Q_j^d + Q_j^s)}$$ Where: $n$ is the order book level $Q_j$ is the size at ...

5

I think one has to distinguish between pricing and fitting/reproducing empirical stock returns. A model might fit the empirical stock returns extremly well but fail to reproduce derivative prices. In my answer I will assume that you are interested in reproducing the empirical stock returns. Mandelbrot and the Stable Paretian Hypothesis The most salient ...

5

Why do you have 16180 observations? Is this daily data over 64 years or higher frequency data? I am guessing so by the magnitude of the intercept. At any rate, your test power would be huge with this large sample size, meaning small relationships will be statistically significant. What Cochrane said is contingent on data frequency. At a high frequency it ...

5

The primary quant skill needed to make the market is optimal control (a typical paper is Guéant, O., L, and J. Fernandez-Tapia (2013, September). Dealing with the inventory risk: a solution to the market making problem. Mathematics and Financial Economics 4 (7), 477-507), because you need to control your inventory and adjust your quotes accordingly: be ...

4

My take on the whole issue is as follows: We cannot be sure to find the one and only true model, the only thing we can do is to identify the most prevalent so called stylized facts and try to model them parsimoniously. The following paper was already mentioned in the comments: Empirical properties of asset returns: stylized facts and statistical issues by ...

3

I can offer an intuitive answer. The limit when your equally weighted portfolio is continuously rebalanced will give you the geometric mean. This is because the excess return of the better performing strategies will be allocated towards the least performing strategies, compounding high returns with low returns.

3

I was going to comment but it turned out to be quite elaborate. My experience with certain AI/ML methods is that they're not deterministic. Take RBM for instance, a very wide-spread paradigm. To train such a machine you have two approaches, backpropagation or Kullback-Leibler divergence. Both require you to initialise the machine to a random state. And ...

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

Definitely time series analysis. What you essentially want to do is some form of impact analysis. this can be done naturally using multivariate time series models like Vector Auto Regression models. Also when working with data to model liquidity you might want to use some specialized procedures like GARCH and ACD. Further there are methods to model non ...

3

There is a major bug when you are getting information from exchanges outside USA. If you get the adjusted prices for BOVESPA (Brazilian Stock Exchange) for example, it will only consider the events that happened using the US Calendar and not the Brazilian calendar of working days, this leads to a lack of information on other exchanges. Be aware of this if ...

3

The main problem with hypothesis testing is going to be survivorship bias. Any manager with a track record you're looking at is only there because they haven't performed badly -- if they perform badly then investors withdraw their money, they collapse, and you don't have their data to do the hypothesis test on! So even if all the managers were investing in ...

3

I guess what they are trying to say here is that, assume you have two time series $X$ and $Y$ which are exactly the same i.e. $X=Y$, the correlation is : $$\rho_{X,Y}= \frac{Cov(X,Y)}{\sigma_X \sigma_Y}\overset{X=Y}{=}\frac{Cov(X,X)}{\sigma_X \sigma_X}=\frac{\sigma_X^2}{\sigma_X^2}=1$$ Now assume a time series $Z=2 \cdot X$, you have: $$\sigma_Z=2 ... 3 Successful strategies in both areas can have the same math requirement. It just depends on the algorithm. PhD level mathematics is not a requirement in either area, despite the impression you may get from academic papers (note that a lot of these papers use math to build a sim market, which is completely dislocated from what a researcher needs to do). I feel ... 3 Unfortunately, the ability and tools to develop a low latency trading system are extremely commoditized and will be insufficient for you to make a living in this field. An overwhelming majority of electronic market makers are staffed 100% by PhDs because trading experience and research compose their primary differentiators, e.g.: SIG EMM - 100% PhD. DRW ... 3 Are you sure the return for two years is 0.7214? It should be 0.3422 per year if you are using 31/12/2011, and 0.3416 if you are using 01/01/2012 as the end date. Assuming the last number (because it makes for two full years, therefore easier to calculate), yes, there is a formula to derive it from the return of the individual years. It's the geometric ... 2 Breakpoint approaches Test based To be well received in a financial econometrics journal, you want test-based approaches. Depending on your question it is common to see a linear regression (least squares) where the parameter suspected of breaking is interacted with an indicator function I(E) where E is the event in question; this function assumes a ... 2 Yahoo data is good enough, but it has its quirks. As people have mentioned, sometimes it does miss out on corporate actions. I remember a while back I was looking at price for Ford (F) around 1999 , and computing my own adjusted close using yahoo's methodology and noticed that yahoo was missing a dividend payment in 1999(which I verified from bloomberg). ... 2 No, it would be$$(RI_{t}-RI_{t-1})/RI_{t-1}

2

The key assumption is that there is no time-series correlation between the error terms. Fama-MacBeth can deal with cross-sectional correlations. See Samuel Thompson's "Simple formulas for standard errors that cluster by both firm and time" in the Journal of Financial Economics (2011) for a treatment of different regression methods for testing equity ...

2

It's probably because of the strong long-standing statistical underpinnings in economics and econometrics, and overall, risk prediction. For example, look at current research with fat-tail distributions and calculations for Expected Tail Loss (ETL), etc. These studies fit Student's t, Normal, Stable, and Pareto probability distributions to data and report ...

2

There are no "fundamental algos" analogous to "technical algos". Instead, quantitative useof fundamental data assumes applying multifactor models to predicting returns and other intrument parameters. That models vary from "academical" (like Fama-French 3-factor or Chen, Roll, Ross) to proprietary models of guys from industry: MSCI Barra, Bloomberg, CSFB, ...

2

Since MiFID 1 in Europe (you may have heard MiFID 2 has been recently adopted), all trades have to be reported somewhere. Reporting channels are numerous: regulated markets have in books and off the book reporting feeds MTF have their own, sometimes one for their Lit book and the other for their Dark one remaining OTC trades have been reported for long on ...

2

In the chapter that deals with NMF of the book "Programming collective intelligence" , the author did NMF on several stock trading volumes and found some comovement. I googled a little. This did NMF on 40 chinese stock close prices. This developed A variant of nonnegative matrix factorization for Stock Trend Extraction. Another google found this also did ...

2

I just ran the Mathematica code below to make a list of stocks and sectors you can access here: https://www.dropbox.com/s/zkzpcnksvfygamp/nyse.xls Mathematica uses curated data from Yahoo. Items that lack sector information have been omitted. The list may be incomplete but it might provide something for you to work with. members = FinancialData["NYSE", ...

2

I do not believe that the exchange is capable of tracking down the person or legal entity who has been part of any recorded transaction and in particular linking the activity of market intermediaries to the ultimate interests of beneficial owners. However, under certain conditions, the person or legal entity has to report to the SEC his identity and his ...

2

On exchanges, there is such orderbook with sufficient amount of limitorders, so when you place an order (market or limit), the "best" limitorders for you will be hitted and change the price last traded price. The price you see is actually just the midpoint between the currently best available bid and ask prices in the orderbook. Therefore, this price might ...

1

Most of what you see on the web is done by amateurs. Many professionals do use 10K and 10Q information in so-called "quant" strategies, but electronic stores of this information don't come free. As a consequence, you don't see a lot on the web about using fundamentals in model-driven investment strategies. Compustat/CapitalIQ is the most well-known ...

1

You are mixing terms here. The definition of an "interest rate" is typically a simple interest rate as applies to the Principal of a loan. The unpaid interest rate is not compounded. This is owed at the conclusion of the loan or when converted to debt. Typically rolled into the equity stake. The definition of "discount" is what the convertible debt holder ...

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