67
Figure 4.10 Trading on Wednesdays.
the S&P 500. The exit is the same as with the bonds shown earlier. Clearly, some are
days better than others
to trade.
Figures 4.8 through
4.12 show the buy signals by day of week;
Figures 4.13 through
4.17 show
sell signals by day of week.
Figure 4.18 shows trading on just the more influential days. The best days to be a buyer were all days
except
Thursday and Friday, while the best sell day was Thursday, with Friday a push, but it is used in the
following listing.
This is not a bad system, it
"made
" $227,822 with 75 percent accuracy on
Figure 4.11 Trading on Thursdays.
70
1,333 trades and had a very small drawdown of only $13,737. I would prefer a larger average profit per
trade than the $170 shown here.
An astute, thinking trader should
be asking questions like,
"Could we use a closer volatility expansion
number to be a buyer on the more bullish days and a farther away entry value on the days that don't work so
well with the 50 percent value? And
how about our exit, would it pay off to hold longer on the more
bullish/bearish days?
"
71
These questions can continue indefinitely, but do need to be asked to optimize performance.
Proof that
research pays off is offered by
Figure 4.19, which shows the use of the preceding rules,
except that the buy
entry comes at 40 percent of the previous day's range added to the open, the sell entry at 200 percent of the
range subtracted from the open.
There is a big difference here; while it actually makes a little less money
($14,000), the accuracy goes to 83 percent, the average profit per trade is escalated to $251, and our number
of trades is reduced by 46 percent!
Separating Buyers from Sellers to Find Volatility
Using Market Swings
A third way to measure potential volatility expansions comes from looking at price swings over the
past several days. Mike Chalek deserves credit for this concept with a system he designed and labeled,
Do'stlaringiz bilan baham: