image image

Voodoo Science-Statistical Analysis

I agree. The unseen data points are the real black swan that cannot be summarized by historical data. Taleb in "The Black Swan" had pointed out that "we focus on pre-selected segments of the seen and generalize from it to the unseen: the error of ...

Voodoo Science-Statistical Analysis

Interesting Study. But I was referring to a different kind of market cycle. image

Voodoo Science-Statistical Analysis

The expected value can be an irrealizable value. A dice roll has an expected value (and a median) of 3.5. Have you ever rolled a 3? The expected value (and the median) of the return of a constant range bar is zero. Unfortunately, even if CRB ...

Voodoo Science-Statistical Analysis

FX ETF Options

Currency safe-heavens, pre-UK referendum

Sometimes(most of the time actually) curve-fitted strategies perform reasonably well under one market cycle. Solid past data? Yes. Can the result be extrapolated? No.

Voodoo Science-Statistical Analysis

I agree. Before applying any form of market analysis one must first perform a logical check first; simplicity & complexity of modeling is not the issue here, the issue is whether a model could stand the test of time a.k.a. robust enough

Voodoo Science-Statistical Analysis

I just see no practical use for it in this case. The median is good to get "a better mean" because it is robust to outliers. A "better mean" leads to more accurate estimation of either location or scale; more accurate estimation of location provides ...

Voodoo Science-Statistical Analysis

Thanks a bunch. I'll study your original post "Auction Market Value Theory & Analytics", no rushy comments before I grasping the rationale behind it.

Voodoo Science-Statistical Analysis

Your rationale behind your claim?

Voodoo Science-Statistical Analysis

/Insert Michael Scott Thank You gif/ It depends on whether you want a robust estimator of location or scale?

Voodoo Science-Statistical Analysis

Haven't used BB for years. The BB I posted above was for demostration only. Nevertheless, a BB with 10 SD is worth a shot. url [/quote]Something way much more problematic than the fat-tail, to me at least, is the high autocorrelation of the price ...

Voodoo Science-Statistical Analysis

Median is supposed to be distribution-free??? Got any papers to back up your claim? Thanks

Voodoo Science-Statistical Analysis

Conventional trading systems based on parametric assumption are capable of generating countless trading signals based on normal price data, collects paper premium till the day of reckoning The opposite system is to trim normal data, but generates ...

Voodoo Science-Statistical Analysis

Theoretical discussions on whether there are better tools for non-normal data analysis

Voodoo Science-Statistical Analysis

Touche

Voodoo Science-Statistical Analysis

After re-reading my own post, yes my point is clear, but allow me to elaborate further 1. Market returns are not normally distributed. 2. Statistics is a form of data science that collects, analyzes and interprets data. Falsely apply parametric ...

Voodoo Science-Statistical Analysis

Let me start with a robust measure of things---Median image Pro: more robust than MA Con: less sensitive & lack of precision

Voodoo Science-Statistical Analysis

Oanda offers variaty of instruments including US bond, gilt and such via mt4. Interactive Broker offers the most comprehensive list of trading intrsuments, stocks, ETFs, futures, bonds, CFDs, etc. But their charting software is...crab.

With which brokers can you trade bonds and interest rates?