|
PureBytes Links
Trading Reference Links
|
Steve,
A few comments:
-My broker has a cutoff time of either 12 noon or 2
pm for entry of mutual fund orders. You may want to check the latest time
for entry of orders. You should also calculate the amount of time youwill
need in order to place the order. Phone orders can take 5-10 minutes
additional. That time should be factored in and may push back the time you
have to make a decision by.
- Backtesting should show stress for the
system. If you are buying, then you should at least buy at the lowest
price of the day, not the 3:45 pm price unless it is the lowest If you are
buying a negative index fund, then you should buy at least at the highest price
of the day. I would also add an additional amount for market movement
after you place your trade. Using a percentage of average true range ATR
would give you a way to do this.
- You might try backtesting with someof the
exchange traded funds. These ETF's trade like stocks, but have the
qualities of mutual funds. They have intraday prices so you can backtest
with SP500 intraday data and execute with the ETF. Some tickers are IWM,
IWN, etc. Of course you would have to get the intraday day for these
funds, but I assume you have intraday data for the S&P, so you should be
able to get data for the ETF's from your same data source.
Regards,
Ed
<BLOCKQUOTE
>
----- Original Message -----
<DIV
>From:
Steve
Davis
To: <A title=amibroker@xxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Thursday, October 31, 2002 2:07
PM
Subject: [amibroker] Backtest: EOD.
Trade: almost EOD.
My mechanical
system generates buy/sell signals using SP500 EOD data and is backtested
assuming the trade is made on the same day before the final closing priceis
known.
<SPAN
class=544235218-31102002>
I am considering
trading the system with mutual funds by obtaining the current SP500
price at 3:45 with the assumption that the prices will not move too much
during the last 15 minutes of the day.
<SPAN
class=544235218-31102002>
<FONT face=Arial
size=2>Does anyone have comments on the
wisdom of this approach?
<FONT face=Arial
size=2>
It would be nice
to somehow backtest the system by injecting noise into the closing prices. I
want to simulate slightly inaccurate closing prices to generate buy/sell
signal, but I must use the actual closing prices for the profit/loss
calculations. I'm not entirely sure how to do this.
<SPAN
class=544235218-31102002>
Any advise or
comments would be appreciated.
<SPAN
class=544235218-31102002>
<SPAN
class=544235218-31102002>Cheers,
<SPAN
class=544235218-31102002>-Steve
<SPAN
class=544235218-31102002>
<SPAN
class=544235218-31102002> Post
AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx (Web page: <A
href="">http://groups.yahoo.com/group/amiquote/messages/)Check
group FAQ at: <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Your use of Yahoo! Groups is subject to the <A
href="">Yahoo! Terms of Service.
|