SERVING THE QUANTITATIVE FINANCE COMMUNITY

Search found 84 matches

by purbani
September 25th, 2013, 11:48 pm
Forum: Numerical Methods Forum
Topic: Historical Database of Macro News with Impact on Financial Markets
Replies: 6
Views: 7335

Historical Database of Macro News with Impact on Financial Markets

Reuters used to have this sort of thing. Google "Event Study" and Reuters or ANO provider. I believe ClariFi have one but have only used the Reuters one which was pretty good back when
by purbani
August 25th, 2013, 5:31 am
Forum: Technical Forum
Topic: Risk measure when you have very few historical data
Replies: 46
Views: 10025

Risk measure when you have very few historical data

Triangular Distribution methods certainly the easiest. A very interesting Maximum Entropy Probability Distribution (MEPD) approach in Chapter 8 of this thesis includes some VBA code to calculate lambda
by purbani
August 11th, 2013, 10:14 pm
Forum: General Forum
Topic: Practical solution for fat-tail risk management?
Replies: 33
Views: 34859

Practical solution for fat-tail risk management?

<t>Hi MizhaelYou may be interested in Why Distributions Matter and;Four Moment Risk DecompositionIn the latter we originally tried and partly succeeded to come up with a 'modified vol' number but the Cornish Fisher expansion cannot model sufficiently high levels of excess skewness and kurtosis to re...
by purbani
August 1st, 2013, 2:12 am
Forum: General Forum
Topic: VaR scenario analysis - does this make sense?
Replies: 3
Views: 7074

VaR scenario analysis - does this make sense?

A helpful dicussion about the differences as well as Component VaR here
by purbani
July 2nd, 2013, 10:54 pm
Forum: General Forum
Topic: nicholas nassim taleb jealous
Replies: 119
Views: 15681

nicholas nassim taleb jealous

An oldie but a goodie - captures some of the pathos and pomposityNNT educates a quant
by purbani
March 22nd, 2013, 9:05 pm
Forum: General Forum
Topic: A hell of a lot to learn from this
Replies: 69
Views: 12181

A hell of a lot to learn from this

<t>JPMorgan Board Says Dimon Should Remain as C.E.O. and Chairman JPMorgan Chase's board said on Friday that it was standing behind Jamie Dimon, the bank's chairman and chief executive, amid calls from some investors that the two jobs be split. In the bank's proxy filing, the 11-member board said th...
by purbani
March 15th, 2013, 11:18 pm
Forum: General Forum
Topic: A hell of a lot to learn from this
Replies: 69
Views: 12181

A hell of a lot to learn from this

<t>Hi HerdYou raise many valid points to be sureIf there was pressure from management to amend the risk numbers downwards and those numbers were presented to any Govt authority then presumably that is fraud. - I am not a lawyerIf the whole purpose of the CIO Office and its location in London was to ...
by purbani
March 15th, 2013, 10:24 pm
Forum: General Forum
Topic: A hell of a lot to learn from this
Replies: 69
Views: 12181

A hell of a lot to learn from this

<t>Jamie Dimon Email Directly Ties JPMorgan CEO To $6.2 Billion Fiasco"On January 23, the risk management group emailed the chief executive to ask for the authority to temporarily raise the bank?s overall risk limit. The same email informed Dimon that the CIO "has developed an improved" risk model t...
by purbani
December 9th, 2012, 11:59 pm
Forum: Programming and Software Forum
Topic: VBA Code for pulling data off Yahoo?
Replies: 12
Views: 170497

VBA Code for pulling data off Yahoo?

Hi AllIf for any reason you need a dynamic price that updates automatically you can use the attached user defined function (UDF) adapted from Randy Harmelink's excellent smf Excel add-in. Note array formula CTRL-SHIFT-ENTER.Kind regards,Peter Urbani
by purbani
September 25th, 2012, 10:09 pm
Forum: Trading Forum
Topic: Singular Spectrum Analysis
Replies: 9
Views: 13768

Singular Spectrum Analysis

<t>A very basic Excel and VBA implementation available Here. The issue about neighbourhood dependence is well made. The non-causality is caused by the diagonal averaging or re-hankelisation in the final reconstruction step which causes re-drawing when new data arrives or the window period is expande...
by purbani
September 25th, 2012, 12:44 am
Forum: Technical Forum
Topic: Curve smoothing
Replies: 6
Views: 22591

Curve smoothing

<t>Singular Spectrum Analysis (SSA) produces a zero lag filter and is easier to apply than Fourier - non-causal though so subject to re-drawing when new information arrives. Can be made causal by taking the end-point or forecast and iterating over some window period. Basic Excel and VBA demo here P....
by purbani
June 28th, 2012, 3:58 am
Forum: Trading Forum
Topic: Copula Based Pairs Trading
Replies: 3
Views: 62945

Copula Based Pairs Trading

There is a little bit about a copula based approach to pair trades at the tail end of this presentation of mine.
by purbani
June 5th, 2012, 10:02 pm
Forum: Technical Forum
Topic: portfolio skewness / kurtosis now with Excel file
Replies: 4
Views: 17411

portfolio skewness / kurtosis now with Excel file

<t>For the sake of completeness herewith Some further revisions to my Excel demo of CoSkewness and CoKurtosis tensor matrices used for 4 Moment Risk Decomposition of Normal and Cornish Fisher ( Modified ) VaR and CVaR. Now includes univariate CoSkewness, CoKurtsosis, CoVolatility and CoDrawdown VBA ...
by purbani
May 24th, 2012, 9:20 pm
Forum: Technical Forum
Topic: portfolio skewness / kurtosis now with Excel file
Replies: 4
Views: 17411

portfolio skewness / kurtosis now with Excel file

<t>Here you go roomer. The spreadsheet now includes the VBA code to calculate the CoSkewness and CoKurtosis tensor matrices which can be used to decompose portfolio risk for the Cornish Fisher (Modified) VaR and CVaR. Please note this is not the most efficient way to calculate these matrices as only...
by purbani
May 2nd, 2012, 10:28 pm
Forum: Technical Forum
Topic: portfolio skewness / kurtosis now with Excel file
Replies: 4
Views: 17411

portfolio skewness / kurtosis now with Excel file

<t>You do not have enough information to be able to calculate the portfolio skewness and kurtosis. Need the time series of returns for the underlying Asset Classes plus the weights which you do have. Currently you only have information about the first two moments ( mean and standard deviation ). The...
GZIP: On