@Autot: Thank you for posting. I acknowledge that this is your first post, so you cannot add screenshots or attachments. I have seen this post by chance. If you wish me to respond, please insert @Fat Tails and I will be notified.
I must admit that I do not know anything about SH index futures. Could you please tell me something about the exchange, the opening hours and where it is located?
The noise bands are based on the volatility of the prior trading days. They show the levels at which breakout moves typically fail, if they are not supported by "other timeframe" traders. Noise traders will usually not push prices outside of these bands.
Of course, the noise bands only show a statistical average. This means that price may revert earlier or later. I personally prefer noise bands over the opening range, because the bands are known with the first tick of the regular session. However, they are not based on the volatility of the current day, but on the average volatility of the prior N1 and N2 days as chosen via indicator settings.
Here some values for the average noise (distance from noise band to regular open) for index futures:
Anyone (even less than 5 posts) can post screenshots. You simply need to attach them, not use a remote image hosting service since you aren't allowed to post external URL's with fewer than 5 posts.
Due to time constraints, please do not PM me if your question can be resolved or answered on the forum.
Need help? 1) Stop changing things. No new indicators, charts, or methods. Be consistent with what is in front of you first. 2) Start a journal and post to it daily with the trades you made to show your strengths and weaknesses. 3) Set goals for yourself to reach daily. Make them about how you trade, not how much money you make. 4) Accept responsibility for your actions. Stop looking elsewhere to explain away poor performance. 5) Where to start as a trader? Watch this webinar and read this thread for hundreds of questions and answers. 6) Help using the forum? Watch this video to learn general tips on using the site.
If you want to support our community, become an Elite Member.
The following user says Thank You to Big Mike for this post:
Thanks for the reply. Apology in advance if any of the terms I use are confusing as I don't know some of the technical terms in English.
I have attached an excel file, in which I have calculated the 3days Pivot Range, Noise Band, Target band for the last 40 trading days. I have also included the Open, High, Low, Close prices of the SHIF in the file for my calculations. It is actually called the Shanghai Shenzhen 300 index futures, but for simplicity let's just call it SHIF....
- In column S to Y, I have highlighted the noise bands which are particularly small
- You can see that the target band values, in most days, are far above the High / far below the Low
Please let me know if I am doing anything wrong there.
Regarding the SHIF, it trades from 9:15 - 11:30, then 1:00 - 3:15 (Beijing Time) whilst the stock market here opens from 9:15 to 11:30, then 1:00 - 3:00.
Note that the 9:15 - 9:30 is bidding time for the stock market and it closes 15min earlier (at 3:00) than the SHIF.
The Excel table basically looks correct. No remarks regarding the noise bands.
However, I am seeing a problem regarding the target bands. For calculating them you need to add an average range calculated over N days to the current low or to substract that average range from the current high. However, for your calculations you use the daily high and low which is not yet known during the day.
The target bands change its value during the day, if you base them on the average daily range. Therefore I have offered a second option to calculate them from the MAX (High-Open, Open-Low). In that case they remain constant over the trading day and you can use an Excel table to calculate them when the market opens (opening price is needed).
The following 2 users say Thank You to Fat Tails for this post:
Amendment : In my last post, the screenshot of the chinese site where it says "month" should have been "week"
Let me detail my method here.
Chart Used : 5min candlestick
Technical indicators used : None
Entry Point: 1). a 5min candlestick opens below the top of Noise Band_Up
2). price go up and break through the top of Noise Band_Up
3). Enter a "Long" position
Stop Gain: At the end of the first 5min candlestick in the opposite direction
Stop Loss: At the bottom of the Noise Band_Up
Note: only 1 trade (buy / sell) will be executed in the same 5 min candlestick period
Attachment Fig.1 shows how a Good day would work out using this method.
S = Short; CG = Stop Gain; L = Long; CL = Stop Loss
The red box marks the opening range of 20mins (9:15-9:35)
Green box is the 3 day pivot range
White box is the noise bands. I only marked it where the price moved around the bands
4 trades would have been executed. (from left to right)
1. Short and then stop gain. Very minimal gain. Position held for 10min
2. Short and then stop gain. Around 7.6pts (=38ticks). held for 20min
3. Long and then stop gain. Around 9.6pts (=48ticks). held for 20min
4. Short and then stop gain. Around 6.8pts profit (=34ticks). held for 10min
To give a clearer picture of the return, at the current index level.
1 contract of index future would require a principle around USD16,600
1 tick (0.2pts) would generate around USD10 (~0.06%)
5 ticks (1pt) would generate around USD100 (~0.3%)
5k - 40k contracts are traded every 5min.
In Fig.2, I show you what a Bad day would look like.
The price kept fluctuating around the Noise Band with no real breakout. This resulted in numerous stop losses.
Assuming I could accurately stop loss at the right price, I would end up losing 10 pts (3% of capital) that day.
The 2 figures above showed when the noise band worked perfectly and when it does not quite work (perhaps due to the close proximity to the 3d pivot range)
Obviously there are things I can add to improve my method and this is what I have been working on.
For example, for fig.2, I could have stop trading in the same direction after 2 failed breakout... but then sometimes the gain you would get from the 3rd breakout (which could be the real breakout) is way more than the 2 losses from the 2 previously failed breakout.
I will add more exceptional cases in my next post.
The following user says Thank You to Autot for this post:
In Column AC & AD, I have calculated the 10d and 20d average of the range.
In Column AE & AF, I have incorporated these average values to the daily High and Low.
I think the way I calculated it is correct, except that the current High and Low are taken as the day goes by (i.e. varying most of the time during the trading period) ?
However, I believe it is fair to say that it should still show some significance in the "Win" days, where the price goes by 1 direction. In this case, whether the current or daily high / low (either one depending on the direction) was taken should not matter.
If I understand correctly, the key to this target band idea is really to calculate how much a move is likely to stop beyond the noise band.
For your alternative option, I guess the way to calculate it is just the same as noise band except that I would now use the MAX value?
I noticed a formula error in my excel file where I used the 10d average range for the Target Band_Up but 20d average range for the Target Band_Down. Although the difference is not big, and the preliminary conclusion I would draw from the set of data I have regarding the target band is unchanged by this.
Following your suggestion, I have then revisited the target band and still find it not as effective as the noise band.
I will dig deeper and see if there is a way to calculate the max gain one could get from a breakout beyond the noise band.
Here comes the exceptional cases.
I would love to improve my method to avoid these from happening but obviously there are always inevitable losses for any trading setup.
I share those to see if you guys have encountered these before and maybe you guys would have thought of a solution already.
- This shows how I would have taken the gain way earlier than the real up move, which comes after a little drop in price close to the Noise Band_Up (but not touching it)
- Maybe the pivot point system can be applied when it breaks through R2, open a "Long" position etc etc
- In hindsight, it looks like it is very logical to long right above the Noise Band_Up when it falls back to it but not touches it. However, when you are actually trading at that moment, it was a place to trade because the next bar could have fallen right through the noise band and serious loss would be taken.
- This shows 1 successful short trade early in the morning
- It is the only day that a very narrow noise band has generated profit
- A large part of the down move is missed out but as I said above, opening a short position at the bottom of the Opening Range is a dangerous move
- The rubber band trade that Mark Fisher mentioned in his book, in my opinion, is not really worth it. Maybe it is my lack of ability to confirm a false breakout but I find that trading that way would suffer in a long period of time.
There are many cases where you profit from a few rubber band trades (e.g. short on top of opening range / pivot range...) but then loses it all in one real breakout.
- shows how a large part of down movement is missed out due to the way I stop gain
- shows a case when the Noise Band is inside the opening range
- a lot of failed breakout, resulting in a few stop losses trades
- This is the worst scenario for stop losses
- Once I longed, normally I should be able to stop loss at the bottom of Noise Band_Up.
This is, I believe, one of the key reason why the noise band is so effective and useful because it is a price level that you know as soon as the market opens. And you know that in most cases (over 80% from my observation for the data I have in hand) that the specific move (could last for 1min - 60min+) that you have traded does not end at the noise band.
In other words, you know for sure that if you long at this point, there is almost always profit coming. it solved one of the 2 key questions in trading (when to enter and when to exit).
- The day shown in this figure has 3 enforced stop losses. By this, I mean when I would not be able to stop losses at the bottom of the Noise Band. The rule I set for this is that if the next 5min-candlestick opens below the noise band, I will close my position immediately and take the hit.
I would have had to do this 3 times on this day (#1, #2 and #4 of the trades), 12.5pt losses
Any comment is welcomed.
It is really an art to setting trading rules, conditions and setup because the more you have, the less "executable" trades you would see in the market yet it does not mean the fewer trades you do will be more profitable.
Who says that the noise bands should be used to trade breakouts? You can either trade a breakout, fade the noise bands or do nothing. It is the larger picture that determines which of the three options you would select.
I would like to see the VWAP and the (1-day) pivot range as well as overbought and oversold levels on the chart.