|
PureBytes Links
Trading Reference Links
|
On Feb 24, 10:50am, Logan467@xxxxxxx wrote:
> Subject: An idea
>
> Here's the idea:
> If the diffrence between the open and close of the Nasdaq is greater than
> the difference between the open and close of the S&P 500 (both are indexes),
> buy the S&P futures either on the close or open. If the S&P outperforms the
> Nasdaq, close out your position.
>
This strategy was, I think, was first recommended by Larry Connors in
"Investment Secrets of Hedge Fund Managers". I just looked in that
book (p. 113). He modifies your approach above, by reversing the
S&P position to short on the open after the a close where the Nasdaq
underperforms. Actually, he uses NDX (the Nasdaq 100) instead of
the Nasdaq Composite. He shows a 56% win percentage, and a
ratio of 1.15 of ave. win to ave. loss, with a max. intraday
drawdown of -20K, over the period 1/20/91 to 11/20/94. Long side
win percentages were 59% and short was 54%. Keep in mind that
1991-92 was up, 1993 sideways, and 1994 was choppy. Still, I was
actually surprised the short side was profitable.
Since Larry doesn't supply the TS code that produced the trade
summary in the book, you have to decide what it means for the
NDX to "outperform" or "underperform". I think you should use
percentage change, but probably both you/Larry were just using the
net number of points difference. NDX is volatile, and sometimes
dramatically over/underperforms the SP, therefore I think it makes
more sense to look at percentage change.
ND futures also are traded these days, though the contract is smaller
than the SP. One thing I looked at a couple of years go, but haven't
followed since was something like the following:
1. calculate the 2-day percent Rate-of-change in the _ratio_ of SP
to ND.
ratio = C of data1 / C of data;
roc2 = 100 * ((ratio / ratio[2]) - 1.0);
2. when the number in #1 above puts in a 1-day swing up (roc2 > roc2[1]
and roc2[1] < roc2), go long SP and short ND, in the proper ratio to
balance their dollar-per-point move.
3. when the number in #1 above puts in a 1-day swing down (roc2 < roc2[1]
and roc2[1] < roc2), go short SP and long ND, in the proper ratio to
balance their dollar-per-point move.
At the time I tried this, I looked at S&P cash and NDX cash, with
an eye on using options on those cash indexes. What I found, if
I remember things right, is this strategy worked pretty well when
the market was moving, but not so well in the flat-line period of 1993.
I think it is also tricky getting the right balance between ND and SP
because ND is much more volatile than SP, and I didn't really try
to solve that.
Larry's got a few other interesting ideas in the same chapter
"New Markets, New Indicators", including a method that uses VIX
(though the levels he uses haven't been seen for several years now),
and an "arb" between the SP day session close and fair value, that
is covered on Globex (based on the idea that SP MOC orders push
the SP futures further away from fair value than usual).
(by the way, you might be able to trade the Nasdaq v. SP spread in
a Rydex fund, which has an OTC fund, and a long/short SP fund. If
fund switches are "free", then there's not much cost reversing
directions. It this works, it can dampen a bit of market volatility.)
--
--
| Gary Funck, Intrepid Technology, gary@xxxxxxxxxxxx, (650) 964-8135
|