Thanks for the suggestion. I know many good and not so expensive sources of tick data as well. The thing is, I like to use the same exact source for backtest and live trading. That way, I don't get any surprise when switching from backtest to live trading. 'anfutures' website look fishy and not very functional at first, but since you've tested it in the past, I might get a shot at them.
Interesting. Any results so far on ETH versus RTH? On average, I think ETH trends better than RTH. This should be no surprise. As for 6 months of data, I agree, but surely it should be enough to give some preliminary results? Maybe I'm just cacthing up with NQ at an interesting time, but I did some more practice trades this morning. Both indices made a major move this morning which was clean on both indices, but ES got stuck after the initial rally moving in a narrow range doing the typical consolidation and back-filling. In the same period NQ moved considerably and is still moving. Daily range so far on NQ: 109.25 points = $2185 per contract. Daily range so far on ES: 22 points = $1100 per contract. You could have entered the same hypothetical trade on both for roughly the same amount risked, so that's twice the juice on NQ for the same amount of risk.
For the exact parameters, I wasn't able to be profitable during ETH. I had to extend the number of period used, meaning waiting more time to enter/exit a position. Trend in ETH seem to be slower to form than during RTH. I'm still validating the ETH because I'm stuck with the 6-month download limitation.
I did a quick comparison this morning between ES and NQ looking at the sum of the daily RTH ranges in points/dollars per contract, # of major swings per day and the sum in points/dollars of the major swings. I only looked at 2023 from start to present day, Friday 26th. I use a proprietary algorithm for calculating the major swings, but basically it captures the major swings for each given day and by design I cap it at 3 swings per day. For example, yesterday had only one major swing (up). Thursday had three major swings (down, up, down) on both indices. As can be seen, the number of swings are pretty much equal with NQ beating ES by 6 swings, so that's negligible. However, when you compare the sum of the ranges and the major swings in $, it's a considerably difference with NQ offering more than 60 % more than ES on both metrics. If you can risk roughly the same per trade (this I don't know for sure) on each instrument, the R/R is no doubt considerably better on NQ. This does not take into account the micro movements within the day where I'm sure NQ will also beat ES hands down offering nice scalping opportunities during periods where ES might be stuck in a narrow range. I may be stating the obvious for the seasoned elite traders on this board.
LF Thanks for your analysis. This is the first time I have seen someone else do this type of analysis and is somewhat similar to what I do. Have you done any analysis that gives you the probability that a major swing will occur during a certain time period in the day and if so the length of the swing. I would think at least 2 of the common major swings are near the open and near the close, but I don’t trade ES or NQ. On may 19 you posted a chart with trades that showed you shorting the last 2 hours of the trading session, so I thought you might have done some probability analysis on major swing times. That was post #42 in “the recent drop in volatility/VIX”. I can see from the chart that you exited on the close and didn’t exit on the 2 prior swing lows. That led me to think that you were “expecting” a larger move down and that you might have also done analysis on how large these “major swings” are and when the probability of a good swing might happen. I daytrade metals, crude and natty. So my analysis starts with breaking up my trading day into 4 hour time periods. Then over some historical number of days, I calculate the average range of each time period like you did for your analysis of RTH. That tells me on average which 4 hour time period has enough range to be tradeable. This only gives the average range and doesn’t tell me if the range was all chop, one single trend, or something in between, so I then look inside each 4 hour time period to determine the probability of 1 or more good trends during that time period and the average move of those good trends. This narrows down which 4 hour time period I have the greatest probability for profit. Cheers toucan
You're welcome. Interesting. Yes, but regarding the timing of the major swings, I haven't found anything illuminating yet, although I'll admit that I can give it plenty more time and study. If you consider the timings of the day low/day high, though, there are statistical significant values to be found. For example, with a day going from low to high, you'll find the LOD within the first hour and a half a great deal of the time. Vice versa to the down side. If I'm not mistaken, I got too greedy/slow on the trade, so my initial entry was early in the day when the market first broke. I added to the trade as I got secondary sell signals on that large reversal higher. As for time, I don't really use it with any deal of precision, but for example for May 19, I knew that if the market didn't trade much beyond where I shorted at that point in time, it was a very small chance that we would see any more upside movement. So, a test of the LOD and a possible break was more likely. But generally, having an idea of where we're at time wise in the session is of course very useful. Simply put: Reversing from LOD to make a new HOD in the last half hour is very unlikely. Have it happened in the past? Yes. Just not very often. So, I like to generally trade with probabilities on my side. Interesting approach. I definitely see some similarities.
Would be interesting if you had any more detailed comments about these markets. Do you trade them using the same base methodology or do you need to utilize different strategies for each market? No equity indices? Thanks.
LF I am a short term daytrader, so most of my trades are less than 10 min and the rest can take 1-2 hours to complete. Across all 5 futures, I see about 10-12 setups a day on a good day and less than half that on a bad day. I trade less than half of those setups. Everything that I say below is based on how I trade and may not be applicable to the way that others trade. I trade metals, crude and natty using the same trading process, from identifying when and where a setup might occur, to the entry process, trade management and exit processes. Some parameters may be different, but everything that I do is based on the risk of the initial stoploss and not ticks or points. So, when using my trading process, all futures tend to look the same to me. For example, if I enter a crude trade, I look at the stoploss as 1R (not points/ticks) and my profit as a multiple of R. that way I can look at any chart or trade and see/use the same process. all of the futures that I trade are affected by “when” American traders start trading each day, the equities open and when European traders stop trading each day. In addition, all 3 metals are affected by economic news and tend to react similarly. often I get into one metal trade and then an able to get into a second metal trade. For metals, I tend to trade aggressively when all 3 are moving in sync and trade less when they diverge. I would guess that is similar when trading NQ and ES. crude and natty are affected the 3 times noted above and are affected when their own reports are released. Economic news doesn’t affect crude/natty as much. I trade 4 hours each day and believe that trading metals, crude and natty during those 4 hours provide the best probability for profits. In the past I have traded index futures and currency futures, but I think that I get the best results for the 5 futures I am currently trading. Some people trade looking to get wealthy, take a lot of risks and usually end up losing. I trade for consistency day-in and day-out. I would be interested in hearing about your comments and methodology trading indexes. I’m always willing to learn and adapt. Cheers Toucan
Thanks for sharing, toucan. Maybe I'll find it worthwhile to look into those commodities at some point in the future. For now, I'm sticking with that I know, and my assumption is that the base methodology I use for ES should be easily adapted to NQ, but I'm doing some testing first of course. Since you mentioned trading hours, would you say your methodology is dependent on time and the reference levels of say the market open and prior day highs and lows and closes? I don't like to go into specifics in public, but my base methodology is based on statistical templates/day types and various statistical biases which helps me to be on the right side of the market. Basically, the idea is to exploit the daily high/low range. Friday was a LowHigh / trend day putting in the low early in the session and the high late in the session. A variation of this day could have seen a larger pullback mid-day, but still with a high print late in the session. This variation is more common than the clean trend day we saw on Friday which is less common. Tuesday was a HighLow / trend day. Statistically, I knew that the best we could hope for was a late session rally to mid range in the last 2 hours. Reversing the day for a new HOD at that point in time would be possible, but very, very unlikely. The rest of the week was different, but all recognizable day types. Simply put, I have many templates/patterns available, so the idea is just to anticipate/predict in advance which it's going to be and trade accordingly. As the day progresses, updated data can be loaded to give more guidance for the rest of the session. Indices move predictably most of the time, but not every day is equal and intraday the market can still throw your a few curveballs. But if one has done plenty of homework and back-testing, there really should be no surprises as every day in the future is a variation of a theme from the past. I got the basic idea from a unique memeber who's no longer active here, but the specifics of all this have been worked out on my own through plenty of trial and testing.