Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
I would like to compare volatility in the market analyzer.
I'm not sure what indicator is good for this.
The reason for example is that something like ATR is really specific to one instrument only.
What's the standard way of comparing the "current action" to historical volatility of an instrument? What indicator is made for that?
in a way that you can then compare a list of instruments and say yeah that one is much bigger than that of the other instrument, so that means it is moving more today - or in the last hour or whatever - depending on how it is configured.
Can you help answer these questions from other members on NexusFi?
This is a fair question. I can see where you are going with this... I don't think there is any indicator or tool that has this ability: "What's the standard way of comparing the "current action" to historical volatility of an instrument? What indicator is made for that?"
This would require the developer of the indicator / tool to know some predefined historical levels and load these into the tool. So this is the part where you will have to do the leg work yourself.
Volatility is easy to measure and quantify and surprisingly easy to classify. But this will fall on you to define. The easiest way to get to a volatility measurement is:
Max(Price) - Min(Price) where the Max and Min functions are taken over a time period you specify. This sample size could be every 10 minutes, every 1000 ticks, etc. You define this. Now the subjective part: Using this definition you have to have a historical baseline for what you think different volatility levels look like: (Low, Medium, High, Crazy, etc.) You will have to define this, but it is easy. For example, on your particular instrument using historical data you may determine the following:
Low Volatility: The market stays inside of a tight 5 tick range
Medium Volatility: The market stays inside of a slightly larger 10 tick range
High Volatility: The market moves in a larger 20 tick range
I don't think there are any shortcuts here if you want to do this right. The math is very simple, but the key points will be to find a sample size of data that is small enough to capture changes, but large enough to see enough movement to be able to have delineation in levels.
Hopefully some of these ideas get you thinking...
Ian
In the analytical world there is no such thing as art, there is only the science you know and the science you don't know. Characterizing the science you don't know as "art" is a fools game.
yup that's the point - I already came up with my own general purpose that does basically what you said with similar caveats - basically a very slow moving instrument will need to use different settings than a very fast moving one - or different chart time frames .. like if it was a bond vs a dax
I thought I read somewhere about a version of ATR used for comparing amongst symbols - however I forget what it was called. And I kind of skimmed that article while not paying full attention so I might have misunderstood exactly what it does
There are a few variables that are pre-defined by instrument such as tick value and average spread size. So these could get loaded into something automatically, which could then compute things....
So the YM is worth 5 bucks and the ES is worth 12.5 per tick, for example.
But short of these kinds of fields, everything else you would need as an input to model this type of information would be historical in nature, and relative in context to ranges, the sample size (time series, tick size, etc.) I can't imagine anyone would ever make something like this and pre-load with their feelings about historical volatility levels. About the best you would get might be a handful of blank fields that users could define their-selves. I think it would be irresponsible for a developer to pick the lines in the sand for users on something like this. From a business perspective it might even come with risks. (I.E their interpretation of high volatility might not align to a users.) But I could way off, and we might find out that there is historical volatility indicator already out there. I just can't imagine it. It almost sounds like magic, if something could do what you are describing.
But if nothing else, you might have just discovered a niche in the indicator space where there is currently a supply gap. If there is enough demand here, someone will eventually develop this.
Best of luck!
Ian
In the analytical world there is no such thing as art, there is only the science you know and the science you don't know. Characterizing the science you don't know as "art" is a fools game.
@iantg Good idea. I will think about offering something for sell.
But, some suggestions for those wishing to do own legwork:
1. Tick value * Range = Dollar Range. You can also do this Tick Value * Range * Volume to create a volume score/money flow.
2. You can use Relative Volume or Relative Volatility. Compare current range to prior range.
3. For something like this, I like to think in statistical terms max, min, average, median. A box-plot or some variation would allow you to gain a better understanding.
4. Other types of relative measures range/vix, range/volume, etc.
5. If you want to compare in standard units, you need to perform a normalization. Z score, otherwise standard normal score, might be worth looking at. You might also try running a Z-score on the ATR.
For the last idea, I pasted Easylanguage snippet, you should able to compare this among instruments and with the other measures such as dollar range, it should provide good insight:
@ycomp The Z-score idea is the easiest to implement. You just need to feed in the ATR as the input (see the EL code I pasted for adaptation). There is free Z-score at link below or check in the downloads section:
@Fat Tails created a few "volatility" indicators for nt 7 and nt 8. some are here in the download section. if you want access to all of them, you can sign up here:
Version 1.3 September 25, 2017
The Z-score or standard score is a numerical measurement of a value’s relationship to the arithemic mean of the data set. The Z-score is the signed number of standard deviations by which the current value of a data point …
However, the z-score does not give you an indication of volatility. It just tells you how far a data point is away from the mean, measured in terms of standard deviations.
There is a statistical definition of volatility which is typically annualized.
To obtain the historical daily volatility you would calculate the standard deviation on daily logarithmic returns over a lookback period of N days and then divide the standard deviation by the square root SQRT(N). The interesting observation here is that volatility is approximately proportional to the square root of time.
For intraday data the definition above is more or less useless, and volatility is typically measured via
- the average range
- the average true range (see Welles Wilder's "New Concepts in Technical Trading Systems")
- the standard deviation
With these indicators volatility can only be correctly measured on time based charts. If you try to measure volatility by applying a range or true range indicator to a range or renko chart, you won't get any new information. Tick charts and volume charts are somewhere between time based and range or renko charts.
Nota: The original average true range (ATR) developed by Welles Wilder uses a Wilder moving average for smoothing. This is an exponential moving average with a smoothing constant of 1/N, where N is the lookback period. In other software packages you will also find other formulas for the average true range. Common variations are an arithmetic average and an exponential moving average that uses the smoothing constant 2/(N+1).
The average true range measures intra-bar volatility. It is only dependent on the size of all bars and the size of any gaps within the lookback period. Otherwise it shows the same result for a trending price action and a range bound price action. The same applies to the average range, which is similar to the average true range, but does not take into account any gaps.
The standard deviation only measures the dispersion the data points used for the calculation. It is not dependent on the size of the bars, but entirely relies on the deviation of the selected input series from its arithmetic mean. Think of the average range, ATR and standard deviation as follows
average range, ATR -> measure intra-bar volatility (volatility inside the price bars, ATR including the gaps)
standard deviation -> measures inter-bar volatility (for example volatility of the bar closes)
Displaying volatility:
If you wish to know, whether the current volatility is high or low, you may simply use the absolute readings of the average range (apply a SMA to the range) or the average true range (ATR). Applying a z-score to the ATR is feasible, but has some limitations. Normally, a z-score makes only sense, if the underlying data points are more or less symmetrically distrbuted around the mean. This is not the case for the ATR, as it may rise ten fold, but not decline further than zero. Accordingly, if you apply the z-score to the ATR, it will never reach the lower threshold of 2 standard deviations, but only helps to identify the volatility peaks, when passing above the second upper deviation.
For stocks and index futures peak volatility is often linked to fear and is followed by upside reversals, so you may use this indicator to detect the temporary sell-offs (chart attached).
