# Analysis of Gaps and Holes- Part 2

**To buy on Gap down or sell?**

The question is not,how frequently there is a gap down[approx 50% of the time of the gaps are down gaps] but what to do, when there is a gap down? To buy or to sell.

The answer is not as simple as the question. To begin with, all I did was use Amibroker to find out, the days when there was a gap down and the day ended positive. Remember, there is a difference between the day ending positive and the gap being filled. How? there can be a gap down of 175 points and the day rallied to end 70 points below previous close.

Another thing, that the data is not detrended. The data range is from October 2005 to April 2009. 4 years of data, around 1100 odd test points, out of which approx 215 days the gap was filled and the day continued to end higher than open.

The motive is: When to buy, if at all to buy, at gap down. Is there any point beyond which the “chances” of the day ending negative[compared to open] significantly reduces?

I will be on most part of the analysis, scaling it as per the 15 day average volatility. That is, if the 15 day average volatility is say, 100 and any gap down of 30 points will be said as gap down 30% or gap of -30%. Savvy?

*Some Statistics:*

Mean Gap Down on days ending positive: -33.69 points

Standard Deviation of gaps on days ending +ve: -8.3

Mean Gap down [volatiliy scaled]:-28%

Std Deviation[volatility scaled]:32%

Although the first part is easily understood, the second part might require some explanation. It just means, the average gap down is around 28% of the 15 day volatility. To be frank, I found the standard deviation of volatility scaled gaps to be quite high.

It can be well discerned that it is not exactly a normal distribution. It has got kurtosis. Without any further, lets check the frequency distribution graph.

The orange line is the visually fitted normal dist. So, I guess the answer is clear from this figure.

Buy on days opening with a gap down of around 20% of the 15 day average volatility, and it has around 60% chances of ending positive[quite an interesting proposition]. You increase your threshold to upto 30% and you increase your chances of the day ending positive by 70%.

So the question, my curious reader, is not then why don’t people use it, but it is, what are the reward to risk ratio of such a strategy. Unfortunately I don’t have the answer right now. Once you start using this as your opening strategy, I would say, you give in yourself to skewed conclusions. Why? We don’t know what is the maximum downside of buying on the opening gap down. Where to place the stop loss, and what are the results of such a stop on the overall profitability of the strategy.

These are important questions and very pertinent to trading as well. In subsequent blog posts, I might talk about it{depends largely on my mood} and try to answer them.

Till then,

Make good trades

*Disclaimers: The *analysis is on NIFTY data.

*Disclosure:* None

*If you liked this post, you might wish to subscibe to Jump Up! for more*