Comments on: Financial Asteroids & Their Effects on Trends

By: Calculating Historical Volatility | Neural Market Trends

Calculating Historical Volatility | Neural Market Trends — Sat, 29 Dec 2007 17:43:00 +0000

[...] IVUnderstanding Fuzzy Trend Following in ExcelBuilding an AI financial market model - Lesson VFinancial Asteroids & Their Effects on TrendsMonte Carlo Simulations For S&P500 Volatility Timing [...]

By: Kerwin

Kerwin — Sat, 19 May 2007 13:51:37 +0000

Looking at trendfollowing in the forex market, i would say we get a financial asteroid at 2 times a week, when the market responds to data about inflation, production, nfp etc. As a trendfollower i have to close out my positions ahead of this data as the trend can either be turned or exacerbated on the data. This puts the time series non-stationary, which makes most backtest using TA null and void during this period of volatility.

By: Tom

Tom — Wed, 16 May 2007 14:06:50 +0000

Mux: The trick is to how to quantify the event into a trend smashing or minor disruptive occurrence. I’m still working on that one!

By: muxControl

muxControl — Wed, 16 May 2007 13:30:07 +0000

So then isn’t it easy to confuse early indication of a financial astriod with the time series changing from trend mode to “non-trend” mode? Because it seems that the onset of both appear to have similar characteristics i.e. correlations become less linear in the transition hence the resulting outlier points.

By: Tom

Tom — Tue, 15 May 2007 21:51:52 +0000

MuxControl: If you would be so kind and point me in the direction of an article about the prediction of the LTCM demise, I’d really appreciate it. Anyone dealing with the amount of leverage LTCM had at its height was asking for trouble, if you ask me.

Some asteroids can be spotted ahead of time, like the insane activity in the Shenzhen. It’ll probably end in tears with some ripple effect worldwide. I don’t know what the extent of the damage would be worldwide but it could turn into a financial asteroid alright and kill a lot trends. Other financial asteroids probably can’t be spotted until its too late.

Time series allows you see the trends, the asteroids hitting them allows you see their demise and ripple effect.

By: muxControl

muxControl — Tue, 15 May 2007 20:58:26 +0000

The demise of LTCM was a disaster waiting to happen. It could have been (and was i think) predicted by some. Lets look at 9/11. Could that and its effects have been predicted? And the recent mass panic in China which rocketed the FTSE, blowing £80bn off its value. Its hard to imagine that the lead up to these type of events can be detected in a time-series. What information would a time-series have regarding the plans of some al-qua-eda guy wanting to blow something up, thereby causing astriod like aftermath?
Can all types of financial astriods be detected?

By: Tom

Tom — Mon, 14 May 2007 20:40:28 +0000

I was introduced to Mandelbrot when I read the Chaos Theory book by Glick and his work at the time focused on weather patterns. I devoured (Mis)Behavior when it came out, fantastic book and its the reason my focus shifted to quantitative and then neural net modeling.

Thanks for the Wilmott link, I’m sure I’ll disappear for a few weeks now!

By: Ashton

Ashton — Mon, 14 May 2007 15:27:12 +0000

I’m a huge fan of Mandelbrot’s work & think that’s one of the best books on the markets. I prefer Taleb’s “Fooled by Randomness” to “The Black Swan” as its focus is more relevant to trading. Highly recommended if you haven’t read it yet. You can find some interesting articles/webcasts by Taleb (& others) on http://www.wilmott.com

By: Tom

Tom — Fri, 11 May 2007 17:54:49 +0000

Ashton: Thanks for stopping by. Yes its those pesky outliers cause a lot of havoc, how often do we just ignore them? Often in data mining we just clip them out of the data as they skew the output. Well the models work great in an equilibrium state but is that reality? The reality is that we will have these “financial asteroids” every so often.

I’ve read a few articles from Mr. Taleb and find him to be interesting and insightful. I plan on reading his new book, “Black Swan.” Have you checked out Mandelbrot’s, “The (Mis)Behavior of Markets?” Its along similar lines that these “asteroids” are part of the overall market organism.

By: Ashton

Ashton — Thu, 10 May 2007 15:16:07 +0000

Interesting post…I think your “financial asteroid” theory is similar to Nassim Taleb’s “Black Swan”. The type of events you’re referring to are outliers, something MPT fails to recognize, which is party why these events have such a great impact. Taleb’s hedge fund has been very successful in trading these types of extreme events. Your quest to uncover financial asteroids may well turn out to be rewarding.