Search
× Search
Tuesday, April 22, 2025

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

Accuracy and smarts first, then speed. Bad data at any speed is still bad data...

photo

 Barry Star, Managing Director, Wall Street Horizon

 Friday, December 12, 2014

That is our mantra everyday...


Print

8 comments on article "Accuracy and smarts first, then speed. Bad data at any speed is still bad data..."

photo

 CV Channagiri, CTO, Head of Operations, Technology Program Manager, Agile Coach

 Saturday, December 13, 2014



Well said.Was it Dijkstra, he renowned computer scientist who said "Early optimization is the root cause of all evils" ... //CV


photo

 Volker Knapp, Consultant bei WealthLab

 Saturday, December 13, 2014



I guess you are, like everyone else here, referring to tick data. You will be amazed how bad EOD data is. Unreliable to say the least.

Optimization is always bad. Parameter stability testing is the term I use.

I have fairly good EOD for SP100, NQ100, DAX and MDAX. If a system performs well in all mentioned markets, then you have something to look at. Did I mention the survivorship bias testing?


photo

 Barry Star, Managing Director, Wall Street Horizon

 Saturday, December 13, 2014



Actually, I'm referring to almost any kind of data. Bad data is bad data... Does it make a difference?


photo

 Kaustabh Ray, Equity Research Technologist and Systems Architect

 Sunday, December 14, 2014



Nice article in the link. Yes smarter is better than faster.

Not only bad data, but also bad algorithm or bad strategy at fast speed spells disaster.


photo

 Barry Star, Managing Director, Wall Street Horizon

 Monday, December 15, 2014



So true..


photo

 Nilanjan Bhowmik, Market Data / HFT Expert

 Monday, December 15, 2014



It depends from the universe you belong to. In my universe both speed and accuracy is important and non negotiable. In fact we consider stale data as bad data.

So let's not put a blanket statement that quality is more important than speed. Both are non negotiable.


photo

 Barry Star, Managing Director, Wall Street Horizon

 Tuesday, December 16, 2014



Data speed and accuracy are inversely related and always negotiable - it's a fundamental law of data gathering. While low speed MAY cause you to lose money, bad data will ALWAYS cause you to lose money. Bad data at any speed is still bad data... Our trading clients know this and that's why they push our firm for accuracy first and speed second (although we do both well). I thought the Tabb article did a good job of presenting this concept as well.


photo

 Michael Storm Jeske, Chief Investment Strategist - III Macro

 Wednesday, December 17, 2014



Even moreso "smart" is much higher scale. HFT seems pretty low scale. Anyone know best revs / P&L per year? Regarding Michael Lewis' book I would've said "Yes, every trade is being gamed and fleeced... For 1-3c per share".

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association