Saturday, April 11, 2009

VWAP, Trend Days and Rangebound Chop Days

Thursday April 9, 2009 which saw a huge 22 point gap-up and rally illustrates a low volume trend day before the Good Friday Holiday (U.S. equity markets closed) where the VWAP (volume weighted average price) served as the ideal first pullback entry point for a decent up-move that lasted through the afternoon. I saved a chart on Thursday highlighting all the good points with the intention of later posting a blog entry. The only thing missing to buy the pullback was a lack of volume confirmation:
VWAP and R2 pivot ideal pullback buy area

On trend-less, range-bound chop days (what famed trader and fund manager Linda B. Raschke calls a "Z-day"), where the VWAP is not well respected (e.g. price will just go right through from both directions) it is still useful because it can give you a clue as to the day's nature. If the VWAP slope is flat and the price crosses through several times that is sort of a big clue. In these sorts of days, some experts (reference to Dr. Brett Steenbarger's article on identifying trend and rangebound days) will fade the VWAP extremes for "reversion to the mean" trades. I am currently using the free NinjaTrader(TM) VWAP indicator available from their support forums (just search the NinjaTrader(TM) Indicator forum for VWAP) that plots the VWAP and is configurable to show up to 3 standard deviations. The thick black solid line on my 10 and 3 minute charts is the VWAP and the thin black dotted bands around it are 1 and 2 standard deviations. I now use these as additional S/R levels and guidelines.

So basically the theory is, on strongly trending days the VWAP itself is used as a retracement area to buy/sell on pullbacks/throwbacks and on rangebound days, the standard deviation bands are used as the buy/sell areas with the VWAP as the profit target area. I'll try to post future blog entries with a link back to this article to see how well this works.

For more on VWAP refer to the popular blog and video series by Brian Shannon