Page 2 of 3

Portfolio envy

Posted: Mon Oct 02, 2006 1:07 pm
by Turbowagon
Hi Stan,

That is a mighty large portfolio you've got there. :wink:

Regards,
Ed

Posted: Tue Oct 03, 2006 5:52 pm
by stancramer
One must be careful to distinguish between "number of active positions today" versus "number of instruments in the portfolio". They are not always the same. For example, I put this little droplet of monitoring code into three different trading systems and ran them on the same large portfolio for 25 years. The plot clearly shows that each system has a different long term average value of %portfolio active.

Posted: Tue Oct 03, 2006 6:23 pm
by Turbowagon
Hi Stan,

Sorry, I must be mis-interpreting the image you sent. It showed apx. 64 positions either long or short. None said out. Now I think maybe you sent an example of active trends and not your actual positions. Is this the case? My question was as to the amount equity you used assuming the 64 open positions. Perhaps, I misunderstood. Thanks.

Best Regards,
Ed

Posted: Wed Oct 04, 2006 5:22 pm
by damian
BARLI - I genuinely do not think you understood my post and most certainly not c.f.' enjoyable response and reasoning, which is a slight yet correct variation of my view and irrespective of that, arrives at the same result.

You could have (but certainly didn't) note that I never said "more sellers than buyers". I said "the number of people with the desire to sell is greater than the number of people with the desire to buy". Desire is the key word.

Taking in a bit of what c.f. said: prices go lower because there are more people with the desire to sell at the last traded price than those with the desire to buy. So at the last traded price, when every bid is filled, there are people left standing on the floor with the desire to sell, yet no one remains with the desire to buy at the last price. These unfilled sellers, plus more then show their offer at the next lower price. And so on. This disequilibrium in desire remains for every tick increment below the last traded price until such a time that equilibrium is found. Prices move down until that point is reached. I have traded plenty of markets with my own account and with the banks account, but I am a small fry. The most amazing bid/offer dynamic I saw was when I sat a few seats away from a guy that traded more contracts in XYZ stock index than any other account, bar none. All day he would ask for a price from the floor and respond with "yours", for quote after quote, he would give every bid size he saw until he found equilibrium. Obviously he didn't do this very often, but when he did: yours!...yours!.... yours! Over and over and over. He pretty much barked "yours!" before the poor floor girl had a chance to give the bid volume over the loudspeaker box, and he'd ask again for the new market before she had a chance to fill the order. It was amazing stuff. Suffice to say within a certain range he had a greater desire to sell at the last traded price (and the next bid below) than the entire market had desire to buy. And so the price went down.

BARLI: I am openly prepared to be wrong in every aspect of life, in fact I very often am wrong and usually I learn from it, but mate, you do not tell me what I do and don't understand (and then back it up, once again, by quoting someone else). Not much irritates me, but that did.

Stan - to answer your quation: I trade weekly bars. Volatility breakout using a very short term measure of volatility (it is obvious but unexpected) all based around a 52 week SMA. I entered near where you did and am profitable on an LB short, several R's. In my account it is the black sheep in recent months. Although long EURJPY is not too shabby.

Posted: Thu Oct 05, 2006 10:15 am
by jankiraly
stancramer wrote:One must be careful to distinguish between "number of active positions today" versus "number of instruments in the portfolio". They are not always the same.
That was a thought provoking plot, thank you! I would say the green system is very different than the blue system; if I wanted to put together a pair of systems with low correlation, green and blue might be an excellent choice. They can't possibly be more than 35% alike, long term, because blue is in only 35% of the same markets at the same time as green. (And sometimes blue might be long while green is short; we can't tell from this plot.)

Red, being in-between, is more similar to green than blue and more similar to blue than green. Red may not be a great choice for diversification purposes.

The last point on the red curve looks like it is about 72%. If 72% of the portfolio equals 64 open positions, that means the portfolio contains about 88 to 90 markets, right?

By the way does anyone have a theory why the green line (dual moving average system) isn't perfectly pegged at 100.00 percent all throughout the test? It it a reversing system after all. I doubt it's because of a code bug, that code looks correct to me.

Posted: Thu Oct 05, 2006 10:54 am
by Tim Arnold
jankiraly wrote:By the way does anyone have a theory why the green line (dual moving average system) isn't perfectly pegged at 100.00 percent all throughout the test? It it a reversing system after all. I doubt it's because of a code bug, that code looks correct to me.
The DMA system has an option to enable Stops. So this will put it out of the market at times.

Posted: Thu Oct 05, 2006 2:24 pm
by stancramer
jankiraly wrote: ... if I wanted to put together a pair of systems with low correlation, green and blue might be an excellent choice. They can't possibly be more than 35% alike, long term, because blue is in only 35% of the same markets at the same time as green. (And sometimes blue might be long while green is short; we can't tell from this plot.)
Jankiraly, I agree completely. I am glad you saw the implications of the three curves! I don't think we have mined all the gold out of this vein just yet.
jankiraly wrote:The last point on the red curve looks like it is about 72%. If 72% of the portfolio equals 64 open positions, that means the portfolio contains about 88 to 90 markets, right?
Absolutely. The PGO system (and the Turtle system, but not the DMA system) has three states: Long, Short, and Out. Some of the markets I trade, happened to be Out on Friday Sept. 22.
.. have a theory why the green line (dual moving average system) isn't perfectly pegged at 100.00 percent all throughout the test? It it a reversing system after all.
There are two possible reasons. The less subtle one was mentioned by Tim A:
Tim wrote:The DMA system has an option to enable Stops. So this will put it out of the market at times.
However, I didn't happen to enable the Stops, parameter settings are shown below. The more subtle reason why it's not 100%, is a lot of fun to discover for yourself. Insert the code above, run the system, and puzzle over the results.

Posted: Mon Oct 09, 2006 12:21 pm
by BARLI
Lumber is limit up today. 8)


Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see

Posted: Mon Oct 09, 2006 1:25 pm
by LeapFrog
BARLI wrote:Lumber is limit up today. 8)


Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see
The trick of course was to get in to lumber Friday which requires a very short term trading approach. I doubt any LTTF would be in lumber yet, but they might be soon if it goes limit up again tomorrow - possible, it did this 2 days in a row last December.

Also, on any long term chart (weekly, monthly) now is a very good time to buy lumber.

Posted: Mon Oct 09, 2006 5:44 pm
by BARLI
LeapFrog wrote:
The trick of course was to get in to lumber Friday which requires a very short term trading approach. I doubt any LTTF would be in lumber yet, but they might be soon if it goes limit up again tomorrow - possible, it did this 2 days in a row last December.
the problem with limit up days is that there are no sellers :wink: those who trade LTTF having their profits erased fast now

Posted: Mon Oct 09, 2006 7:39 pm
by billpritjr
BARLI wrote:Lumber is limit up today. 8)


Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see
Wheat is at 10-year highs.

http://www.bloomberg.com/apps/news?pid= ... ommodities

We have chart patterns and fundamentals to confirm the uptrend. What is the trendfollower to do?

:)

I am short the Yen also, chart reasons (new 8 month lows), and fundamentals, Japanese economy, North Korea, etc.

Was it Seykota that said new highs and new lows are signifigant?

Posted: Mon Oct 09, 2006 9:04 pm
by edward kim
BARLI wrote:Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see
yes, and i like it.

your prediction is gutsy.

in general:
this structure means up this much.
that structure means no more.
but never this structure means up this much and no more.

i like to see wheat lock limit up again tomorrow.

Posted: Mon Oct 09, 2006 9:36 pm
by billpritjr
edward kim wrote:
BARLI wrote:Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see
yes, and i like it.

your prediction is gutsy.

in general:
this structure means up this much.
that structure means no more.
but never this structure means up this much and no more.

i like to see wheat lock limit up again tomorrow.
Dennis or Eckhardt?

Posted: Tue Oct 10, 2006 3:48 am
by edward kim
billpritjr wrote:Dennis or Eckhardt?
both.

and also statistical significance tests show this behavior.

Posted: Tue Oct 10, 2006 6:24 am
by BARLI
Edward, good research! Wheat is already limit up in the night session of ECBOT, I'm wondering if Corn and Soy will follow the fashion

Posted: Tue Oct 10, 2006 9:26 am
by Old European
What contracts I'm holding? Curious?

I'm long Wheat and KC Wheat!

And furthermore:

long positions:

British Pound
T-bond, 10- and 5-yr T-note
Nasdaq
Nikkei
Rough Rice

short positions:

Japanese Yen
Swiss Franc
Crude Oil
Natural Gas
Lean Hogs
Lumber (since many many months)
Cotton
Sugar

But I'm not at all excited about it. I just trade my system and I'm detached from the daily market volatility. I furthermore sincerely think that the above information is pretty useless without any knowledge of the mechanical system I'm using, the composition of my trading universe and my investment time frame.

Cheers,

Old European

Posted: Tue Oct 10, 2006 1:52 pm
by edward kim
BARLI wrote:Anyone's trading Wheat? I am biased its going to 500 cents and then sells off, but we'll see
yes, and i like it.

your prediction is gutsy.

in general:
this structure means up this much.
that structure means no more.
but never this structure means up this much and no more.

i like to see wheat lock limit up again tomorrow.

Posted: Tue Oct 10, 2006 4:32 pm
by MarkH
I thought I would join Old European in posting current position information. I agree with his assessment that it is mostly useless even though it may be interesting. It tells you nothing about why and when the position was entered or why and when it will be exited. These are the net positions from an overall program trading four separate strategies -- some of which may be long while others are may be short (or flat). For example, I am overall net short the yen, even though one system/strategy is long. I have not had much luck including wheat in my research -- however, a very reasonalbe (if not, compelling) case can be made that it (and other historically poor performing liquid markets) should nevertheless be included to deal with one of the most troublesome forms of curve-fitting/optimization -- the choice of markets traded. Alas, maybe wheat is getting ready to prove this!


Long:

corn
soybeans
10-yr note
bonds
dollar index
s&p

short:

coffee
crude
cattle
hogs
natural gas
swiss franc
yen
sugar
cotton
nasdaq
lumber

flat:
aussie dollar
russell 2000
dow

Posted: Wed Oct 11, 2006 3:14 am
by AFJ Garner
"So perhaps this thread should have been called something along the lines of "What positions are you holding and why?" "

Posted: Wed Oct 11, 2006 5:15 pm
by Dean Hoffman
Coming into today

Long KC Wheat
Long MW Wheat

Short Cotton

Long Euro Stox 50 Index
Long IBEX 35 mini Index
Long Hang Seng mini Index

Still holding all on the close but all stops pretty tight.