Jeff from System Trader Success published an interesting trading idea that is based on the difference between the low and the closing price of the bars on a candlestick chart. He noticed that almost all of the bars on a Euro chart he was analyzing had pronounced tails. This led him to investigate whether or not he could build a quantitative strategy from this idea.
From the very beginning, you can sense that he isn’t sure this will be a good idea:
This got me thinking if I could buy near the daily low I might just have an edge to capture those ticks. And by the look at some of those tails, there is a lot of dollars just sitting there.
But what I’m talking about is bottom-picking. How often are we told not to pick a bottom or picking a bottom is a fool’s game? Such conventional wisdom may be true. I don’t know since I’ve not seen or performed such a study.
Once again I remind myself, in the trading world what passes for conventional wisdom is often nothing more than a myth.
Jeff starts by formulating a plan to test the validity of his Low-To-Close (LTC) concept:
The LTC observation is the price difference between the closing price and the low of a single daily candle. The first thing I want to explore is how many ticks are available between the low and the close?
To do this I created an EasyLanguage strategy that computes this value for each candle and places them within an Excel spreadsheet. Within Excel I can compute the median ticks for each day and generate a histogram showing the distribution of ticks.
He then tests his concept on the Euro futures contract from August 8, 2008 1月を通じて 27, 2011:
This study produces 1,061 days with a median LTC value of 66 ダニ. This value translates into $825.
Below is the distribution curve which shows a nice smooth graph which peeks at bin number 3 あります。 30-44 ダニ. The curve smoothly tapers off into a nice tail as the bin numbers climb.
Notice the spike in bin 21. This demonstrates the “fat-tail” phenomena often associated with trading.
He then introduces a trend filter and compares trades that are with the trend versus trades that are against the trend:
I was surprised to discover the median number of ticks between the With Trend (64 ダニ) と Against Trend (70 ダニ) studies were so similar.
This suggests there just might be equal opportunity in both bullish and bearish markets.
Looking at the distribution curve for these two studies does seem to show a left-handed shift in the curve when we are against the major market trend. Maybe being more aggressive in profit taking would be warranted in this market climate.
He concludes by suggesting a couple of ideas that could lead to developing a system out of this concept and seems to believe that this idea is worth further analysis:
In the end it sure looks like there might be potential here. The trading concept would be to open a long position near the intraday low and closing that trade at the median LTC value or at the end-of-day.
The vast majority of days close 15 ticks or more above the daily low. That’s a lot of ticks that can be captured!