Very Interesting, as always jack, very good post. I fully agree with you, and you explained it 100 times better then I, and put it in a better perspective for me..lol. I would think skill level definetly would be the determining factor. That and the type of strategy, I assume, relative obviously to liquidity. (Which loops back to skill.) - secXces
Do many scalpers use automation? does this remove skill? do these automators need to concentrate their time to find new "cans" to put the market in constantly? Michael B.
This interests me aswell. This thread has me intrigued..lol Im sitting here refreshing the page over and over waiting for reply's..lol - secXces
i am afraid there is a real clearcut answer to which style is best. first of all you need a measure. i would propose any kind of risk/return ratio that takes smoothness of the equity curve over time into account. easiest is sharpe ratio. now consider a long term trendfollower and all his trades over ten years and assume he trades a sharpe of 1. move on and consider you could have exactly the same curve within one day. before you had a daily sharpe of 1, meaning that you will experience huge drawdowns of 20% or more regularily. now, when you are trading a sharpe of 1 on minute data within a single day, you will hardly ever experience a down day. the same sharpe on minute data (namely 1) results in a huge sharpe increase on daily data. now comes the problem: you cannot easily replicate the daily sharpe of 1 on minute data without influencing the market very quickly. so the amount of money you can take out of the market is much smaller than with the daily system, yet it is much more stable. that is the whole game about it, i am afraid. everything else is people talking their own book, most of the time because they face constraints which do not allow them to go to higher frequencies. and many discussions are not clear in what is compared with what. if a low frequency trader trades twenty systems at the same time, his results will be more robust against market changes. compared to a high freq guy who exploits one small edge that could vanish overnight.
man: good point (I used to trade multiple stocks but I still think it's inferior). Also there are types of edge that never change (eg structure) and there are fads (like Level II in a blow-off bull market) that go from rags to riches to rags.
Good point Man. In economic terms its the law of decreasing marginal utilities that catches up to the scalper quite quickly, entering part 5 of the utility curve where he starts losing money as he adds more volume. It doesnt take too much volume per trade for the scalper to affect the market, since he does so many trades per day, his accumulated volume per day grows exponentially. Now the question is, would the scalper and the large position trader both reach part 5 of the utility curve at the same level of daily volume? Or would one of them reach it first? If they both reach such a level at the same volume, or if the position trader reaches such a level before the scalper, we can say that scalping is a more profitable method. However, if the scalper reaches that level first we have to ask ourselfs, is the difference in volume against the scalper enough to undo the edge he has from operating in smaller waves? Another factor to consider is the alpha exponential of the utility curve for scalping, or said in friendlier words, the rate at which scale economics affect the scalper as he starts working with other scalpers. To answer this question he should ask ourselves, would 10 scalpers working together trading 10 different simbols gain 10 times more than a single scalper trading a single symbol, or would it be 11 or 9 times more? and would the result be consistent if we had 100 scalpers trading 100 different symbols? By answering both this questions we would be able to determine if scalping is indeed the most profitable method for operating the market, or if it is the most profitable method for operating up to a certain volume where other methods become more efficient, and if so, what method would bring us closer to the long term equilibrium?
Not much if you use IB or another per share comm system. I scalp equities so with IB .01 of any move goes to commison up to 2000 shares.
OP, When I saw the title of this thread I figured I was in for some good entertainment. At the end of this mess I realise it has killed more brain cells than last nights fall off the wagon. Thanks, I think...