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)
Same comment here. No problem with the VWAP itself. But the standard deviation bands are too narrow for a trending day. I suppose that the chart created with SierraChart shows 2 SD bands.
If the VWAP and the SD bands are correctly calculated, they should have exactly the same value on any type of chart. After all, a VWAP is the volume weighted average price of all trades over a specified lookback period. The standard deviation is also calculated from all trades over the lookback period.
Therefore neither VWAP nor standard deviation bands depend on the chart type and should always be identical, whether displayed on a minute chart, a volume chart or a fancy Renko chart. This is not the case for the Sierra Chart indicator and clearly shows that the calculations are false.
Let us go one step ahead and explain why Bollinger Bands to do not work correctly, and why most of the charting packages use a slow formula to compute them.
(1) The first flaw of Bollinger Bands is that they only use the bar close as an input. Other data which is available, such as the open, high & low is neglected. The important point here is that each price bar stands for a price volume distribution, which is unknown. The market profile theory has the price bar divided into time price opportunities (TPOs). Basically this is the same as assuming a linear intra-bar distribution of volume. The next refinement was to take into account the total volume of the bars, which resulted in volume-weighted time price opportunities. However, the linear price volume distribution neglects the fact that the aggregate volume per price level is typically higher within the body of the bar when compared to the upper and lower shadows. This problem can be solved by calculating the standard deviation from high, low, open, close, midline and midline of the body.
(2) The second flaw of Bollinger Bands is that neither the moving average nor the standard deviation take into account volume. Replace the SMA with the volume-weighted moving average and replace the simple standard deviation with a volume-weighted standard deviation, and everything is good.
(3) The formula used for computing the standard deviation typically calculates the difference between contract price and the mean, then adds up all the squares of the differences and divides by the number of contracts to find the variance. The standard deviation is finally obtained by extracting the root from the variance. The problem is that this formula can be painful slow for larger samples. A simple updating formula can be used, but is widely unknown.
To fix the three problems, I have written a simple indicator - called standard deviation bands - that deals with the problems.
Daily chart: The upper chart shows the correct standard deviation bands. Statistically, about 95% of all prices should lie within the bands. The bands should also act as support and resistance. If you look at the original Bollinger Bands, you will notice that they do not open up as quickly as they should during a strong trend, and that they open up too much on the wrong side. There are clearly more than 5% of all prices outside the bands. Support and resistance do not match.
Intraday chart: The disparity in volume, for example between regular session and pre-session is much higher for intraday bars compared to daily chart. For that reason the behaviour of Bollinger Bands is even worse, when applied to intraday charts.
Why I have posted this here? The standard deviation bands calculated for the VWAPs are suffering from the same three problems:
-> only the close is used, although information on open, high, low and close is available
-> volume is not taken into account
-> instead of a fast updating formula, mostly the traditional formula is used to calculate the standard deviation
Do your comments/views pertain to just Ninja Trader or have you run some BB on Sierra Charts and verified that they suffer from the same limitations/weaknesses?
The Bollinger Bands use a well-known formula. By definition they all implementations of the Bollinger Bands indicator suffer from these weaknesses. I have basically suggested to replace SMA / StdDev in the Bollinger Bands indicator with a VWMA / volume-weighted standard deviation and to use all data available from price bars, not just the close.
I do not know which formula for the standard deviation is used by the different charting packages (see point 3 above). NinjaTrader uses a slow formula, and my suggestion was to use an updating formula instead, which would reduce CPU load for longer lookback periods. The updating formula can be used for both the classic standard deviation and the volume-weighted standard deviation indicator.