Years mon tue wed thu fri 2006 -1% 0% 8% 1% -2% 2005 4% -3% 1% -2% 0% 2004 1% 6% 4% -4% -3% 2003 7% 9% -5% 7% 7% 2002 -9% -15% 17% -4% -5% 2001 -3% 4% -4% 11% -13% 2000 20% 5% -18% 4% -15% This is just the beginning of my 75 years file, which is attached here below. On the last row I set the formula to an average of years 1988-2006, but it can be changed to an average of 1930 to 2006, or to a SUM of year x to year y. I got the data for my sums of weekdays' gains from Yahoo Finance, but I couldn't post all the data, because it won't allow the upload here, nor would it work for many of you because I have my excel set on European date format.
The problem when you say that a strategy works well over a period of 10 years is that you will have to wait at least 10 years to see it it worked or not...
Does it mean we can buy the dips on Monday and Tuesday, hold it into closing and be profitable in the long run (at least theoretically)?
Yes, it does. If what's happened in the last 20 years will keep on happening, then you would make money doing what you just said. Yet, it would make more sense to open long on friday at close, and close long on wednesday at close (and viceversa). Just theoretically though, because I haven't tested such a system, nor know anything else about it, like drawdown and such. I was trying to use these data for my own system, but they don't seem strong enough to be used. I don't know, I am still pondering.