SPY - Some Backtesting
Q: I had some time this morning, and ran a quick backtest on SPY. I chose the high, low, close divided by 3 and looked for 4 down days, like last Friday. Then, the following day bought at the market and sell at the close. Since 1993, it showed 100% return on your money doing 176 trades. The gambling odds at buying the next day after;
A:
I've got a much fuller set of statistics on this, and virtually every
strategy you could think of; you know, a 10-bar histogram of the
distribution of results, simulated compounded trading results, etc.
The bottom line though, in so many cases just reduces to the following
question: Can the strategy beat buy and hold?
For the very
most every broad market index or average or representative
portfolio of stocks or index shares is that buy and hold beats
most TA techniques of all kinds. The more strategy evaluation
that you do, the more you will find this to be true.
This is because the long-term averages, and the "average"
stock, tend to have a long-term upward drift in price that
compounds naturally over time. TA techniques generally work
best--for a while--on stocks and averages that go nowhere but
up and down over time.
I said many times that reading the group is my "daily affirmation".
Sometimes I worry that people will smarten up and start
trading intelligently, reducing my profit potential. Then I read this
group, and see many $millions in my future.
Yeah, I used to have a pretty strong emotional component to
trading. I would get more charged up than when shooting craps
in Vegas, probably because there was so much more money
involved.
As I increasingly let most of my trading fall under computer control,
some of that emotion is gone. On the other hand, the amount of
money just keeps getting larger, and there is a certain thrill in that,
plus the practical consideration that I'm gonna need it cuz I don't
think my 401Ks or Social Security are gonna be worth spit in
a few years.