High frequency only works if you have an edge and you have adequate capital. I will not go into the details of my edge. You have assumed that I do not recognize certain logic. Thats fine. Youre view does not influence my work thank God. Now for those who may want to seriously think about he subject. Here is an item to consider. Systems with an edge periodically lose it. Why? Scientist, you are well ahead of me, surely you know the answer. Its basic parametric statistics. I'll wait here. No hurry.
Yeah I've understood that you're in the market on trend days. So am I. And what do you the rest of the time? Wasting your time! Let's face it, that's what you're doing if you're sitting in front of the screens 80% of days, not using any opportunity to trade: Wasting precious time. You might as well be playing golf. As for trading the ranges with indicators (like you said in the following post), well, I won't abuse you about this, but will give you a friendly hint: Indicators don't work. If you want to learn to trade ranges, orient yourself on the bars, i.e. define the ranges, S/R levels, volume, even pivots. Just forget the Stochastics etc. It won't work... As for using the insulting term "moving-average-crossover-trader" earlier in this thread, based on your trend trading, I deduct that you actually might be one - in that case, I hope you don't take any personal offense, and hope you do well with whatever you're doing! Best Regards, S
Absolutely, they do. That's why you periodically re-adjust your parameters. The principles, however, remain. S
Why would I be sitting in front of the screen on non-trending days? Not me sir. I try to have a life. Actually moving averages are indicators that actually do add value to trading. Crossovers however do not work well, except in trending markets like currencies. I don't use them in the ES market, because there is no need. Like you, I have a diagnostic that does not rely on traditional indicators. Once the early market passes my test, I simply enter once. From that point on, I am managing size and waiting for either a drawdown that exceeds my limt or the close, whichever comes first. You may not like to acknowledge it but we probably manage our positions in a similar way. I take profits periodically (when the market tells me that I am wrong) and I do so my looking at an envelope to signal my early exits. Thank you, Lefty.
Reasonable answer. My view is different. What I have found is that price exhibits random behavior and then transitions into periods of non-random behavior (stationarity). There is a way to analyze this action so that you can determine when that is going to happen (approximately). Currently, the temporal window is about 32-38 bars (periods). Once you know that you have a choice to make. Optimize, wait for the next "non random" period or use a different system. Most traders don't have the background or inclination to test, so they simply trade several systems concurrently. They view this as one system hedging the other. Thats fine, except that again, you are exposed to increased risk for about the same return. By the way, your concept of high frequency trading allowing your edge to work is correct (I agree) in theory. However if your edge changes for the reasons I have suggested, high frequency trading turns into a double edged sword. It is a good thing then, that you know how to test for system failure. Nice talking to you. Lefty.
<I>You are welcome to do your own maths on it, and you will see I'm right.</I> No. Wavetrader is right and you are agreeing with him LOL I didnât say anything about frequency. A tight stop does not necessarily increase your frequency. However you are correct in that a higher frequency compounds profits faster. I believe that is true to the point where transaction costs become excessive. I havenât done the math. If you pull your stop in you will want to make sure your system trades more often. You may also want to pull your profit target in to see if your frequency increases and allows you to compound faster. Higher frequency doesnât happen automatically. Your expectancy score will tell you where you stand. I have only backtested the strategy I posted the link to earlier, never traded it. I have a feeling there is an execution factor involved (slippage and missed fills) and I donât know exactly how it might affect returns if at all. Scientist please share your experience.
That's why you continuously evaluate your trading cycle. If your trading cycle requirements aren't met 100%, you immediately know to stop and re-evaluate your system. Regarding the "trading cycle": It is a common systems evaluation tool. I wrote about this some time ago, but later found someone else had written about it here already, and since you might not deem me as credible, you might give more credibility to that person's definition, since it's actually acrary... http://www.elitetrader.com/vb/showthread.php?s=&postid=159224 Read the bottom post, it explains everything. As you can see, there is no room for ambiguity as to whether your system is performing or not. Enjoy! S
That is an asinine comment, and I've made this point long before WT even strolled these boards. But all the power to you, as you wish. I agree, tighter stops don't necessarily increase your frequency. However, they can, if your parameters recur frequently. I.e. the quicker you exit a trade, the quicker your system is "receptive" for a new trade, thus increasing frequency. But I need not go in on that, I'm sure you understand the dynamics. I do very well, because I tend to backtest most things by hand. Also, yes, transaction costs are a factor. You will have to either limit your frequency to an acceptable level relative to costs, or look to lower your transaction costs (better clearer, exchange seat). If you do higher frequency, it could be very justified, and give you the edge. As for slippage and missed fills; When testing a system, I always account for slippage, and count NSF's (not-sure-fills) as non-fills, i.e. no trade. That is the only way to know with certainty. If, after all this, your system still has positive expectancy, then real-time results should be just as good or even better. Mind you, I would not accept slippage for any kind of entry, only exits. For entries, I use limit orders. Use stop exit slippage accounting for your trade size relative to the typical depth in that market. I have a special spreadsheet that calculates the probable slippage for me, based on the size I'm using. Hope this helps, S
I come along and stir up a hornets nest and you guys are back on systems. Why the hell can't you just look at the chart and see what the pro's are looking at? You are doomed if you don't. BBG