[R-SIG-Finance] What's are some go-to packages in R/Finance for detecting shocks in financial time series?

Alexey Zemnitskiy alexzemnitskiy at gmail.com
Tue Sep 29 02:16:02 CEST 2015


It should, we tested the package for illiquid stocks with 1-2 trades a day.

The models would require sufficient warm-up period. Since this is not a HF
dataset, I would also suggest switching off market microstructure noise
detection
<https://www.portfolioeffect.com/docs/platform/quant/manuals/portfolio-settings/model-pipeline#noise_model>
-  the estimated HF noise at EOD frequency is close to zero for liquid
stocks anyway.

If you want to give intraday data a try -  SPY, GOOG & C sample history is
available through the package.

Best,

Alex



2015-09-28 19:26 GMT-04:00 Ilya Kipnis <ilya.kipnis at gmail.com>:

> Alexey,
>
> As someone who works independently, unfortunately I don't have access to
> intraday data at all. Does that work with daily-frequency data, like the
> kind I use on my blog?
>
> -Ilya
>
> On Mon, Sep 28, 2015 at 7:24 PM, Alexey Zemnitskiy <
> alexzemnitskiy at gmail.com> wrote:
>
>> Hi Ilya,
>>
>> If your focus is on intraday price shocks - you could check out our
>> PortfolioEffectHFT package.
>>
>> Right now it's available at:
>> https://www.portfolioeffect.com/docs/platform/quant/downloads
>> It would also be available on CRAN shortly under BSD license - we are
>> doing second round of submission corrections.
>>
>> The setting you might be interested is
>>
>> https://www.portfolioeffect.com/docs/platform/quant/manuals/portfolio-settings/model-pipeline#jumps_model
>>
>> It is using a combination of several jump detection methods (quantile,
>> wavelet-based, etc.).
>> For intraday volatility estimators see portfolio_variance
>> <https://www.portfolioeffect.com/docs/platform/quant/functions/absolute-risk-measures/portfolio-variance>
>> & position_
>> <https://www.portfolioeffect.com/docs/platform/quant/functions/absolute-risk-measures/position-variance>
>> variance
>> <https://www.portfolioeffect.com/docs/platform/quant/functions/absolute-risk-measures/position-variance>
>> methods.
>>
>> PortfolioEffect service is free to use with your own pricing data.
>> There is optional access to HF market data history for 8000+ US equities
>> since 2013 if you need that.
>>
>> Best,
>>
>> Alex
>>
>>
>> 2015-09-28 17:55 GMT-04:00 Ilya Kipnis <ilya.kipnis at gmail.com>:
>>
>>> So, I'm back to researching trading strategies on volatility. However, as
>>> the mailing list knows, volatility ETFs are characterized by price shocks
>>> more often than not, causing rapid drawdowns. One example would be, say,
>>> the closing price of XIV from late April to mid-May in 2010, late 2011,
>>> the
>>> SPY correction in 2011, or the more recent one last month during the
>>> China
>>> meltdown.
>>>
>>> Does anyone have any R package that they can recommend for detecting such
>>> quick corrections in a systematic manner?
>>>
>>> Thanks a lot.
>>>
>>> -Ilya
>>>
>>>         [[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-SIG-Finance at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
>>> -- Subscriber-posting only. If you want to post, subscribe first.
>>> -- Also note that this is not the r-help list where general R questions
>>> should go.
>>>
>>
>>
>

	[[alternative HTML version deleted]]



More information about the R-SIG-Finance mailing list