Trading Pearls

Quote from nonlinear5:

I had been developing automated trading strategies, and my approach was essentially this: if certain thresholds were exceeded, the strategy would take a long position, and if the same threshold were exceeded on the other side of the spectrum, the strategy would take a short position. The same thing with the exits. That is, the underlying assumption was that the market movements (up and down) were symmetrical with respect to my conditions. In the more commonly understood terms, think of a simple MA crossover strategy. When the fast MA is above the slow MA, you buy, and when it's the opposite, you sell short. Same thing here, the assumption is that the market moves symmetrically up and down in response to these crossovers. In reality, it's well understood that the price symmetry is not there. The down moves are generally faster than the up moves. There are numerous implications deriving from this fact, and one of them is that the duration of the short trades (that is, the time while in a given trade) can be much smaller than the duration of the long trades, while you realize comparable reward/risk ratios. Thus, the entries and the exits (and even the indicators) have to be treated differently depending on whether the strategy is trading from the long or short size. Ultimately, I separated my strategies into the "short only" strategies and the "long only" strategies, and everything fell in place. They are running totally independently, using their own indicators, conditions, entries, and exits.
More...

For me, entries are the same. It's stop movement and exits that differ.
 
Somebody asked how to identify trend days.

http://www.barchart.com/chart.php?s...ddindicator=&submitted=1&fpage=&txtDate=#jump

Have a look at TRIN.

If you have an extreme reading Greater than 3 or less than 0.5 and it does not correct at the opening then you are likely looking at a trend day.

For instance the first day in this chart shows serious selling at the open which does not correct. Interpretation: Program selling.

The middle (or second day) shows a sinusoidal pattern within the extremes rotating about the axis of 1.0: Interpretation range day.

Yesterday (last day on the chart) You had extreme selling which immediately corrected. Interpretation Fade the gap and a countertrend day at least to fill the gap.

There are other ways to identify program trend days but this is an easy method.
 
Quote from Zr1Trader:

I use two 23 inch monitors with 6 markets on each screen. Then from there I can tell by blending candlesticks ect if I'm getting a signal on a different timeframe. My signals are pretty rare but effective so that is the reason I monitor so many different markets.

ZR1 does imply the vette, I don't have one yet but will eventually.

Your vette sounds awesome, that is almost exactly what I would do to it.

At the moment I have a 94 z28 camaro ony 50k miles on it yet. Got it for 4k from a police auction, It was an undercover narcotics car (b4c) steal of a deal. I put some things in it. Headers, Custom true dual exhaust, vette wheels and tires 17", shot of NOS in case I feel like racing some big boys. Saving for that vette!
More...

I'm sure when that NOS kicks in, it's a "kick in the seat!" Pretty wicked stuff.
 
Quote from Lojanica:

Somebody asked how to identify trend days.

http://www.barchart.com/chart.php?s...ddindicator=&submitted=1&fpage=&txtDate=#jump

Have a look at TRIN.

If you have an extreme reading Greater than 3 or less than 0.5 and it does not correct at the opening then you are likely looking at a trend day.

For instance the first day in this chart shows serious selling at the open which does not correct. Interpretation: Program selling.

The middle (or second day) shows a sinusoidal pattern within the extremes rotating about the axis of 1.0: Interpretation range day.

Yesterday (last day on the chart) You had extreme selling which immediately corrected. Interpretation Fade the gap and a countertrend day at least to fill the gap.

There are other ways to identify program trend days but this is an easy method.
More...

Interesting. I extensively tested TRIN (with a bar chart feed) a few years ago, both with the precise rules as outlined in John Carter's Trade the Markets or whatever its called, as well as with various other rules that I created. In the end, i cancelled my bar chart subscription because TRIN was no better than a random distribution graph.

Maybe i'll give your rules a look. Predicting trend days would indeed be the holy grail.

I generally accept that nothing available (meaning advertised or distributed online or published in books) is profitable. That's not how the system works.
 
Quote from IronFist:


I generally accept that nothing available (meaning advertised or distributed online or published in books) is profitable. That's not how the system works. [/B]
More...

One possible exception might be a system so stupid as to defy usage.

It's the clever stuff that sucks.
 
Quote from IronFist:

Interesting. I extensively tested TRIN (with a bar chart feed) a few years ago, both with the precise rules as outlined in John Carter's Trade the Markets or whatever its called, as well as with various other rules that I created. In the end, i cancelled my bar chart subscription because TRIN was no better than a random distribution graph.

Maybe i'll give your rules a look. Predicting trend days would indeed be the holy grail.

I generally accept that nothing available (meaning advertised or distributed online or published in books) is profitable. That's not how the system works.
More...

I don't have a set threshold, that is 2.5-3 for program selling and 0.5 for program buying. The key is the value coming to the baseline where a balance exists, which is 1.0. If the buying/selling doesn't "correct" to the mean then you might be looking at a trend day. There are other rules like the slope and direction is important not necessarily which side of 1.o you're on but Wednesday's action and associated $trin showed all the characteristics of a trend day.

Premium can also shed light on these days as do support levels. If price is acting funny at support then again the players who support at support are no longer supportive of supporting the supporters. If you know what I mean.....
 
Quote from jhiro:

hwjnsx.jpg
[/QUOTE

http://upload.wikimedia.org/wikipedia/commons/6/65/Pearlofallahsm.jpg

Now this is a pearl.
More...
 
Quote from IronFist:

Interesting. I extensively tested TRIN (with a bar chart feed) a few years ago, both with the precise rules as outlined in John Carter's Trade the Markets or whatever its called, as well as with various other rules that I created. In the end, i cancelled my bar chart subscription because TRIN was no better than a random distribution graph.

Maybe i'll give your rules a look. Predicting trend days would indeed be the holy grail.

I generally accept that nothing available (meaning advertised or distributed online or published in books) is profitable. That's not how the system works.
More...

TRIN (The ARMS Index) kept me outta trouble insofar as my predisposition was to buy. But with TRIN rising-----the index is inverse to price---buying would have been a bad move. In fact once we were above 1.5 we were solidly into "trend Day territory given we never corrected back the the "balanced buying and selling" of 1.0

In the end it was a trend day down. VERY DIFFICULT to predict but this is a start.

Cheers all.
 
Quote from Lojanica:

http://www.barchart.com/chart.php?s...ddindicator=&submitted=1&fpage=&txtDate=#jump

I suck at teaching. But if anyone is interested have a look at the TRIN charts at the Barchart URL.

The key is direction, rotation about 1.0, and the magnitude of the buying or selling.
More...
Do not ask me how it works or why but I suppose it's because it shows what the underlying trend for the day is. Like anything it is subject to failure and whipsaws but with a little filter you could trade off this chart alone just fine.
 
You must log in or register to reply here.
Back
Top