[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[amibroker] Re: Longterm equity chart (3 possible cuases for Jan 2001 jump)



PureBytes Links

Trading Reference Links

Chuck,

For simplicity and understanding why not just post the .png of the 
equity curve generated by PT, i.e. the first two columns of the 
account file preferably plotted on a log scale

Fred

--- In amibroker@xxxxxxxxxxxxxxx, "Chuck Rademacher" 
<chuck_rademacher@x> wrote:
> Wow, b....
> 
> Granted, I just got up, but I'm overwhelmed by your reply.  I 
haven't even
> had my first cuppa yet!!
> 
> There are no "penny" stocks, unless you call a stock that traded 
above $1 a
> penny stock.
> 
> I would have thought that the trade listing that I posted for DT 
answered
> all of your questions, but I will endeavour to explain.
> 
> The equity curve is actually a "profit" curve.   I used the 
word "equity"
> because everyone else does.   It shows the cummulative profit/loss 
for each
> trade.
> 
> I generate this profit curve by:
> 
> 1.  Setting the initial capital in the account for each stock to 
500,000 (an
> inflated figure that doesn't come into play).
> 2.  The system trades 50,000 per trade and doesn't worry about 
running out
> of money.
> 3.  The syntax for my composite is:   AddToComposite (Equity()-
500000,
> "~Composite", "X", 7);
> 
> There is no cummulative effect of thousands of stocks being in the 
basket.
> The profit/loss from each trade is added/subtracted and then the 
stock is
> forgotten.
> 
> The system uses the VLIC for timing only.  It does not trade the 
VLIC.   The
> system certainly does not just buy all stocks when the trend is 
up.   There
> is system logic that I have not supplied.   I have said that it 
uses an
> "oversold" indicator for each stock.   A stock must at least $1 
(actual
> price) on the day, can't be above $15  and must have a reasonable 
average
> turnover in order to be considered.   Then, it must also 
be "oversold".
> You may see prices in the trade listing that are outside of the $1 
to $15
> range.  Remember, that the filter is based on the actual price on 
the day
> and the trade is being calculated based on the backadjusted price.
> 
> There is additional logic that decides that the signal is only 
valid for a
> few days.   It has a "use by date".
> 
> When the market timing signal reverses (goes down), individual 
stocks are
> sold if they are going down.   If an individual stock continues to 
go up, it
> is held until it starts to go down.
> 
> This is why yoiu will see stocks being bought a few (secret?) days 
after the
> trend turns up and sold a few (secret?) days after the trend turns 
down.
> 
> Other reasons for a stock being sold are the 40% stop loss or if it 
ceases
> trading.   Remember, my database has extinct stocks.   These show 
up in the
> trade listing with symbols like ~UNIV, ~LAC, etc.
> 
> The trade listing shows that the system bought 124 stocks based on 
the buy
> signal that was generated on 26 December, 2000.  It started buying 
on the
> 27th.
> 
> You can see that some more stocks were bought over the next few 
days.   This
> is part of the logic of my trading system, not the timing system.   
Pretty
> soon, I will have specified the entire system in my positings.
> 
> The exit signal was generated on the 8th of February, 2001 and the 
system
> started dumping the entire basket on the 9th.   It took a few days 
to unload
> the entire basket, again based on the system logic.
> 
> I hope that I have covered all of your issues, but I'm happy to 
delve
> further.
>   -----Original Message-----
>   From: b519b [mailto:b519b@x...]
>   Sent: Friday, June 20, 2003 7:50 AM
>   To: amibroker@xxxxxxxxxxxxxxx
>   Subject: [amibroker] Re: Longterm equity chart (3 possible cuases 
for Jan
> 2001 jump)
> 
> 
>   Chuck,
> 
>   Your explanation for the huge spike in Jan 2001 makes sense - 
that a
>   there was larger number than normal of stocks that qualified as
>   being beaten down - likely due to tax loss selling in Dec 2000.
> 
>   However, this explanation (or my interpretation of it) raises a
>   question about the equity curve's significance.
> 
>   I see three possible explanations. I think the third is the most
>   likely, but I want to mention the other two first so you can 
confirm
>   whether or not they are factors.
> 
>   HYPOTHESIS ONE
>   Your system began to trade penny stocks in late Dec 2000 or early
>   Jan 2001. Penny stocks can have periods of huge gains of 70% or 
more
>   (and can have similar losses!)
> 
>   Although I expect your system has a minimum price that excludes
>   penny stocks, I thought I better ask just in case.
> 
>   HYPOTHESIS TWO
>   Let's say by Dec 2000, the system has generated a cumulative total
>   of 1,000 stocks it had traded. This number is important if I
>   understand the way the equity curve is generated. Because your
>   inactive stocks are padded out to the present for the sake of the
>   equity curve, that implies that once a stock gets added to the
>   equity curve, it remains part of that curve to the end, even if it
>   never trades again.
> 
>   So, the jump in the equity curve in Jan 2001 could simply being 
the
>   fact that a disporportionate number of new stocks qualified for
>   being part of the equity curve for the first time. If 700 new 
stocks
>   qualified, that would account for the jump of 70% even if those 
new
>   stocks were not profitable.
> 
>   Or perhaps (I would guess probably) you have already compensated 
for
>   this by subtracting the initial equity give to a stock before that
>   stock's own equity curve to the cumulative equity curve, or 
should I
>   say "profit" curve. Something makes me think that you probably 
have
>   done this compensation step. If so, we need to go to hypothesis
>   three.
> 
>   HYPOTHESIS THREE
>   In anticipation of a reply that the chart is a "profit" report, 
what
>   I think has caused the jump in January 2001 is that a large number
>   of new stocks has added a large amount of modest profit all at 
once.
> 
>   I am up early this morning (about 4 hours of shut eye) so my ideas
>   as well as words may be a bit fuzzy. Perhaps an illustration will
>   help.
> 
>   Let's suppose that the "profit" curve is based on the profit of
>   about 1,000 stocks by Dec 2000. Let's also assume the total
>   accumulated profit is about $4,000,000 by Dec 2000. Let's also
>   assume that there was a much larger number than normal of new 
stocks
>   meeting the beaten down requirement of your system - perhaps due 
to
>   tax loss selling in December. Let's say 1,000 new stocks got added
>   to the stocks already part of the profit curve. Assuming the that
>   Jan 2001 entry (maybe it was a very late 2000 entry - same 
result),
>   was moderately profitable with an average gain of $3,000 for each
>   $50,000 position (that would be same as the average $3,000 you
>   report). Those 1,000 new stocks times 3,000 would add $3,000,000 
to
>   the profit curve.
> 
>   Thus, the jump in the profit curve of 70% could be caused by the
>   total number of stocks being traded increasing by 70% even though
>   they only made a modest, average profit on that trade signal. Is
>   this hypothesis, or a variation of it, possible?
> 
>   b
> 
>   --- In amibroker@xxxxxxxxxxxxxxx, "Chuck Rademacher"
>   <chuck_rademacher@x> wrote:
>   > VLIC is the Value Line Geometric Index.   I only came across it 
by
>   accident,
>   > but it makes an excellent overall market timing vehicle.  If you
>   care to run
>   > a 4/11 EMA cross on it, you will see the exact turning points 
that
>   I used.
>   > Data for it is commercially available starting in January, 1985.
>   I have
>   > created my own values for it back to 1970.
>   >
>   > I think that I explained why the system did so well in January,
>   2001.   I
>   > believe that many stocks were oversold in December, 2000 and 
this
>   system
>   > picked took advantage of that situation.   Such spectacular
>   (unbelievable?)
>   > performance is only by circumstance.   The system was not curve-
>   fit to take
>   > advantage of that situation.  In fact, I wish that the huge 
spike
>   wasn't
>   > there.
> 
> 
>         Yahoo! Groups Sponsor
>               ADVERTISEMENT
> 
> 
> 
> 
>   Send BUG REPORTS to bugs@xxxx
>   Send SUGGESTIONS to suggest@xxxx
>   -----------------------------------------
>   Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
>   (Web page: http://groups.yahoo.com/group/amiquote/messages/)
>   --------------------------------------------
>   Check group FAQ at:
> http://groups.yahoo.com/group/amibroker/files/groupfaq.html
> 
>   Your use of Yahoo! Groups is subject to the Yahoo! Terms of 
Service.


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/ySSFAA/GHeqlB/TM
---------------------------------------------------------------------~->

Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/