10/20/2017 FRIDAY : Hot Topic

# Where is the market likely to close?

### Calculating the close

This article assumes a basic understanding of Market Profile and the terms used in Market Profile. No attempt is made here to explain those terms.

### Open Trade Close Market Profile

The purpose of this study is to determine where the market is likely to close and trade given one of three Market Profile positional openings. The market can open above, in or below the Value Area. We want to know by the end of this study (1) where the market is likely to close (given its open) and (2) where the market is likely to trade (given its open).

The types of questions that you will be able to answer at the end of this article and studying the tables in it are:

• If the market opens above the Value Area then what is the chance that it will close (1) above the Value Area, (2) in the Value Area, and (3) below the Value Area? (and all other combinations of this question)
• If the market opens below the Value Area then what is the chance of it trading (at some point during the day) in and/or above the Value Area? (and all other combinations of this question)

This article does not attempt to suggest or give trading recommendations as no calculations have been done on Value Area size and/or distance the market may open from the Value Area. The data calculated for this article was calculated with the intention of producing tables of probabilities which will eventually lead to back tests being run using Market Profile to help develop strategies.

### Fair Data

The first question I ask myself when back testing a strategy is: What was the back drop to the period in question when the strategy was run? If, for example, our strategy was to buy the open and sell the close and our back test showed this to be a fantastically profitable strategy I would be less impressed if I was then told that the period over which the test was run saw the market rise by 50%. I would also want to know how many days of data went into the back test.

It's no different for creating statistical tables. You need sufficient data over a period which has seen a good cross section of market conditions.

The data set that I've chosen to create these tables from is the E-mini S&P500 (the ES) between the dates of 1 April 1998 and 27 October 2004. Between these 2 dates the ES changed from 1120.00 to 1124.75 which is +4.75 points or +0.4%. During this period the market gained a high of 1574.25 (+40.6% off start date) and a low of 767.25 (-31.5% off start date). I feel that you can't really get much better sample dates than those with the start and end value so incredibly close to each other and a significant high and low over the 6.5 year period. The sample also includes one major price shock (New York terrorist attacks on 11 September 2001) and many minor price shocks.

### Open to Close

The following shows a cross tabulation of where the market closed given where it opened. Before looking at the result table I wrote down what I expected the results to be. I expected that if the market opened above the VA then it is more likely to close above the VA than in any other area etc. These are the results of around 1,700 trading days of the ES.

Close
Above In Below Total
Above 65% 19% 16% 34%
Open In 35% 33% 32% 36%
Below 14% 21% 64% 30%
Total 39% 25% 36% 100%

The top row shows us that of all the trades that opened above the VA, 65% closed above the VA, 19% in the VA and 16% below the VA. The next two rows are self explanatory and follow the same pattern. These figures are pretty much what we would expect except that the middle row is slightly off but by little enough that we can ignore this.

The bottom row (Total) shows us that irrespective of where the market opened (i.e. ignoring where the market opened) the market closed above the VA 39% of the time, in the VA 25% of the time and below the VA 36% of the time. This is more skewed than I would have initially expected.

The Total column at the end shows the averages of where the market opens in relation to the VA. This is roughly split into thirds each which is what we would expect.

On subsequent inspection the closing averages (bottom row) are not as surprising as I would expect. The value area is a restricted area based on 70% of the previous session's TPO count. As such the area above and below the VA gives far more area to move in and therefore more opportunity for price to occupy. i.e. more data points which the price can trade at exist on either side of the VA (far more prices) and this I believe explains this apparent anomaly.

This next table shows us the percentage of times the market traded above, in or below the VA given where it opened.