Discussion in 'Automated Trading' started by Craig66, Nov 6, 2012.
----Seeing an offer did not guarantee that I could buy it. By the time my buy order got to the exchange it was very possible that that offer would have been cancelled.----
i've been saying this for many years,yet many people still disagree..HFT is good and providing liquidity,right?
Nobody who understands the markets is disagreeing with you. They're just saying you need to get faster.
If you are using IB TWS, you will probably always miss.
in these days by the time my order reached the exchange everyone and his mother would know what i was doing . SEC is helping too. by demanding to check margins etc..that would give everyone,but me another reason and excuse to delay my order.
Yup, run your own brokerage and you can do those non-risk tests internally. As an individual, that's just one more roadblock in your already poor execution speed.
This whole article appears a little fishy. He curve fits the hell out of 10 second price action, and he talks about how much time he spent making a perfect simulator. Then no mention about transition to live trading and says he made 500K with almost no drawdown, but profitability only lasted a year. But altogether he says development only took 6 months. Prior to that he sold 2 businesses for a mere few hundred K each to fund this endeavor. Then after the profitability is lost, he tries to sell it unsuccessfully (this is believable) then hangs up the code and gives up. Most of these statements seem a bit hard to believe.
well the year was 2009 and volatility in the market was high. he traded fdax and russel futures
I find them to be entirely believable, because they are similar to my own experiences.
It's really, really hard to get edges to stick for a while. Also, the whole 2009 and start of 2010 thing is telling, because the market today is absolutely nothing like those markets. I have a closet full of edge-skeletons who died of starvation after 2010.
okay two things that are believable: profitability didn't last (of course, it's curve-fit), and he couldn't sell the algo. other than that, the rest of the story is still fishy.
Some comments are very interesting ...
Separate names with a comma.