Price movements in the same direction?

kkseal

Well-Known Member
#22
I think its a gud idea overall, however if one is simply interested in price persistence, a simple runs test will do. I dont think one can find any meaningful/significant probability relationships with the above-mentioned approach.How do I know? Been there, done that, got the trophy :D

Other concepts like gaps might be more meaningful.
CV, how about a thread from you on testing? I mean the different methods, what's suitable for what, which ones can be automated, which ones are better to do manually - that kind of things.

Recently saw a beautiful island reversal pattern on the Sensex charts (a clustered one hosting a triple bottom) with gaps on either side. The one on the right got filled three days later; but sometimes such breakaway gaps remain unfilled for weeks, months (even years maybe).

What to test about gaps? The probability of a gap Up/Dn opening getting filled intraday?

Regards,
Kalyan.
 

oxusmorouz

Well-Known Member
#26
Case 1 review : Change in the closing price of one day is not a sigificant indicator of next day's more.

Case II:
Objective: To determine if the direction of open and close (C > O or C < O) of one day acts as an indication for the direction of price movement the next day.

Data used:
Dow Jones Industrial Average from 1970 - Daily data. Total bars = 8,241 bars of data.

Condition:
If Yesterday's Close > Yesterday's Open and Today's close > Today's open (or) if yesterday's close < yesterday's open and today's close < today's open, plot 1, otherwise plot 0.
The sum of the above is added and divided by the total sample size (bars used).

Probable inferences:
a: If the probability is significantly higher than 50%, then, today's price direction can be used as a tool for tomorrow's price in the same direction as today (with open and close as the base).

b: If the probability is significantly lower than 50%, then, today's price direction can be used as a tool for tomorrow's price in the opposite direction as today. (with open and close as the base).

c: If the calculated probability is neither significantly higher than 50% or significantly lower than 50%, indifference exists.

I use 50% mark here because that is the probability through coin toss that C > O or C < O.

Inference:
From 1970-2002, there existed a probability of 51.1467% that price will move in the same direction it did the previous day. Putting it in other terms, 48.8853% of the time, it moved in the opposite direction.
I have once again not validated this statement by using statistical hypothesis testing but by view of the eye, the conclusion that can be reached is that historically, today's relationship (direction wise) between open and close does not serve as a sufficient information as to tomorrow's price movement.

I have attached the test result.
I request members to comment on this.

Obliged.
 
Last edited:

winstonn

Well-Known Member
#27
Hey Oxy,

You are doing a great job!
well, i suggest u to read ' Definitive Guide to Futures Trading ' by Larry Williams.

he's also done lot of research like you and on variety of markets including commodities, stocks, Bonds, etc.

~Winston:)
 

kkseal

Well-Known Member
#28
Inference:
From 1970-2002, there existed a probability of 51.1467% that price will move in the same direction it did the previous day. Putting it in other terms, 48.8853% of the time, it moved in the opposite direction.
To quote Thomas Alva Edison

'Now we know another thing that doesn't work'

(Edison had to make hundreds of experiments before that electric bulb finally lit up. Everytime an experiment failed to produce the desired result, the great man would smile & make the above comment)

Regards,
Kalyan.
 

oxusmorouz

Well-Known Member
#29
To quote Thomas Alva Edison

'Now we know another thing that doesn't work'

(Edison had to make hundreds of experiments before that electric bulb finally lit up. Everytime an experiment failed to produce the desired result, the great man would smile & make the above comment)

Regards,
Kalyan.
That's precisely why this thread was created. There already exists a few threads on strategy development which focus on "what to do?" but not "what not to do?" leading to the case where people develop good strategies from rules with no edge (as in the case with "Drawing good conclusions from erroneous assumptions").
Those who are new to development of strategy like myself can skip a few steps on what should not be done, if satisfied with this report, rather than spend time experimenting again.
If however, we are fortunate to find some tool with some probabilistic edge, we can use it ourselves :)
Irrespective of the case, we learn quite a deal through first hand analysis, don't we?
 
Last edited: