The Personal Financier: Roller Coaster Ride in a Manic Depressive Stock Market: Researching the Option to Make a Quick Profit

I’ve decided on objectively researching the possibility of making a quick profit off the recent ups and downs. Here are the results

In the past couple of weeks I’ve written quite a bit on the manic-depressive behavior of the stock market and the possibility of making a quick, yet risky, profit from it. As I’ve described in length I avoided gambling on this phenomenon for various reasons.

I attributed my “success” at guessing the market’s direction in the day to come to various biases and luck. Usually when I do decide to make a short term move I more often lose money than make a profit.

Numbers are usually reliable and I turned to them to take an objective look at the phenomenon. Markets with high volatility (or “fear”) often act in a manic-depressive pattern sinking to new lows daily usually followed by a short rapid daily incline. I decided to test this out to see if it’s based on more than my feelings.

I’ve analyzed the S&P500’s behavior in the last 20 years and tried to make sense of the manic-depressive market. What I searched for is short term market behavior which suits what I described. In short: I searched for days with unusual negative or positive returns to see what happens in the following day.

I think we’ve all noticed how September presented us with several examples of daily negative returns of -4% and up which were followed by days with significant positive returns (only to return to the negative a day later).

My findings were quite interesting

My findings were quite interesting. The total number of trading days I analyzed was 5,230 in the past 20 years. In order to substantiate my findings I had to fine-tune “unusual” return. Therefore, in my analysis I looked for days in which the negative returns were lower than -2%, -3% and -4%. Then I conducted the same analysis for positive days in which returns were higher than 2%, 3% and 4%.

My first finding was very interesting itself. The numbers of days on each side of the spectrum (positive and negative) were amazingly quite the same. Out of 5,230 days of trading in the past 20 years:

144 days generated negative returns lower than -2% vs. 143 days generating returns higher than 2%.

35 days generated negative returns lower than -3% vs. 39 days generating returns higher than 3%.

11 days generated negative returns lower than -4% vs. 14 days generating returns higher than 4%.

I love symmetry and the symmetry revealed on each side of the return spectrum is pretty interesting.

My second finding shows days which generated negative returns are more likely to be followed by days with positive returns. Trading days that generated positive returns have an almost 50% chance of being followed by another positive trading days. Trading days that generated negative returns have a much higher chance of producing a positive return the next trading day.

My third and most important finding shows that in accordance to our intuition the higher the impact in a certain day the higher the chance the next day will produce an opposite result.

A trading day which followed a trading day which generated returns lower than 2% had a 63% chance of generating a positive return.

A trading day which followed a trading day which generated returns lower than 3% had a 71% chance of generating a positive return.

A trading day which followed a trading day which generated returns lower than 4% had a whopping 91% chance of generating a positive return.

As aforementioned days which generated positive returns display a more averaged behavior:

A trading day which followed a trading day which generated returns higher than 2% had a 45% chance of generating a negative return.

A trading day which followed a trading day which generated returns higher than 3% had a 51% chance of generating a negative return.

A trading day which followed a trading day which generated returns higher than 4% had a 64% chance of generating a negative return.

What can be said really?

It appears the results are in accordance with our initial intuition although some results were more surprising than others.

An annoying paradox or built in flaw of this sort of research is that the rare phenomenon we’re looking for are, in fact, rare. As such any statistical induction would be highly inaccurate. It seems the pattern is in accordance with out intuition but it is suffice to look at yesterday to throw us back into blunder.

Yesterday was a negative trading day (even though it stated higher than +2%) which followed a pretty nasty trading day in which the S&P500 took a -3% dive.

The problem is that we, as investors, never invest in the average but in a specific day, period, portfolio, etc. Still I think this was a rather interesting experiment.

Related posts:

Image by: gwburke2001