trend following delusion shattered

Status
Not open for further replies.
Quote from Mathemagician:

Could one say that they're betting on a trend in the spread?

M
More...

Well, in practice I think they most often are just using fundamental analysis. I could be wrong, but I don't often hear of M&A arb players doing things like only buying as the spread contracts and shorting it if it begins to tighten. If anything, a pure arb player would to the reverse, eg see any unwinding as a "buy the dips" opportunity to get in at even better prices.

They very well may unwind their positions, but for fundamental, not technical, reasons. Ie maybe the government says they won't approve the deal any more. In this case they will unwind their position as it no longer makes fundamental sense.

I'm talking just about arb trading in its pure form. This isn't to say some people don't use a combination of "arb" and "trend following", nor am I saying trend following doesn't work (nor that it does work lol though I believe it does in some cases). I'm just saying that doing arb trading in its pure form is very different than trend following.

-Taric
 
Quote from Thunderdog:

I actually anticipated your response. Your point is valid.

The method I use identifes minute directional bias a varying number of times during the day. Based on nothing other than historical testing and ongoing real-time observation while trading, I use an artificial, minimal "expected" (for lack of a better term) amount of movement based on a homemade calculation of very short-term volatility. It is a crutch that I use, and a benchmark at which time I take partial profits and move my protective stop to break-even for the remainder of the position. This crutch has no predictive value because sometimes the price does not reach it and I find myself with a loss. At other times, it may exceed that benchmark several-fold. It is simply an artificial mechanism that I use to exploit my assessment of a momentary directional bias.

I suppose you can make the argument that this initial benchmark is an expected minimum value of sorts. But it is a best-guess bet rather than a prediction or a legitimate "expected value."
More...
If it helps you trade, that's great.
 
Quote from ptunic:

I actually think both sides are right on this question.

I absolutely agree that on a high level arbitrage is NOT related to trend trading. And the way most people play arbs in practice, there is nothing to do with trends.

For example, take simple M&A arbs. Oracle is buying company XYZ; you simply use fundamentals based on the deal offered as well as the % likelihood of it occurring and you come up with the most likely price for the newly merged company. So you short Oracle and buy XYZ. This has *nothing* to do with trends.. people who play this particular arb could absolutely care less whether Oracle is hitting new highs or lows. They are simply betting on the spread between the two companies as they converge into a single business entity. (Actually this example isn't truly an arb since there is some theoretical uncertainty as to the true % chance of the merger being successful, so the examples of cash versus futures are probably a better one).

That said, I think Lefty is correct as well. He clearly wasn't talking about the high level conceptual aspects; rather he was talking about a very specific cause-effect relationship, namely the impact of M&A arbs on lower-time frame trends, ie an external higher-time frame event and its impact as far as making lower-time frame trends more predictable.

So again I think both are right but you are sort of talking about different things or at least in a different context a little bit.

-Taric
More...


great post. thanks....i agree with you. however, the little leaguers trend definition based on "index arb" was bizzare to to say the least--- cause and effect is not a defintion.
 
Quote from 5yrtrader:

Ok thats why Seykota, Henry, Eckhardt, Dunn and countless others run billions of dollars of investor money using trendfollowing models.

I have yet to see a scapler running any sort of fund. But I am sure that you can give me a reason why. Maybe you are going to buy the Red Sox from your profits scalping.
More...



there are many funds that trade without a trend. in fact the number one fund over the last three years over 100 million is managed by a non "trend believer". the traders you mention as super succesful trend followers are merely the outliers in the distribution curve--- many ( perhaps thousands) have failed trying to trade the trend.
 
Quote from NickelScalper:

Setting d is not about a price target. It's the minimum expected price differential between prospective entry and exit, which a trader should come to some sort of conclusion about before making a move.

That is what d is in the challenge question, which I repeat for purposes of reference:

What is a reliable method that can be applied to past price action to determine a usefully large positive or negative number d such that (p1-p0)>=d, where p0 is the current market price and p1 will be the market price in the near future?
More...
NickelScalper, just to get some clarity because I see a lot of people getting hung up about d. Is it correct that you are not asking for the value d, but for a method of predicting future price action such that a profitable trade can be taken (where d is minimum price move that is "worth" trading)? I would imagine that d depends on the spread of the vehicle being traded and the round-trip commision cost.
Is that correct or are you asking for the value d as part of this challenge?
 
Quote from winter:

NickelScalper, just to get some clarity because I see a lot of people getting hung up about d. Is it correct that you are not asking for the value d, but for a method of predicting future price action such that a profitable trade can be taken (where d is minimum price move that is "worth" trading)? I would imagine that d depends on the spread of the vehicle being traded and the round-trip commision cost.

...
More...
That is correct.
 
Quote from winter:

NickelScalper, just to get some clarity because I see a lot of people getting hung up about d. Is it correct that you are not asking for the value d, but for a method of predicting future price action such that a profitable trade can be taken (where d is minimum price move that is "worth" trading)? I would imagine that d depends on the spread of the vehicle being traded and the round-trip commision cost.
Is that correct or are you asking for the value d as part of this challenge?
More...

Nickel:

I already pointed out this is done by a mechanical trader using backtested results in the form of average trade. You want an exact expected result per signal; this is not how trend following works. You take the signal under the expectation that, on average after commissions and slippage, you're going to net x dollars. This is "d". It is constant through time, and once the system is not delivering "d" (determined through simple statistics) you reassess it or can it.

Also, I did some research on your past. It seems you're a failed "daytrading trend follower":
http://www.elitetrader.com/vb/showthread.php?s=&threadid=43533&perpage=6&pagenumber=4

You scalp the QQQ, but why would you do that when you can scalp the NQ for MUCH cheaper costs? And why would a scalper be paying for trading signals?

I think you're simply a struggling trader. There is nothing wrong with this, but I also don't think your position allows you to be so critical of one's personal approach to trading if you yourself are still new and struggling.

You've recieved answers that follow your specifications several times, but you dismiss them by slightly changing the rules each time. Your rules are in opposition to the whole method of trend following. A trend system that relies on profit targets exclusively is usually not as effective as one that doesn't (this doens't mean they shouldn't be used; they are good for scaling out and whle leaving a portion to run for the trend duration). You are essentially looking for a system where the exit is only on the stop or target, and at no point in between.

That is ineffective trend following (in the majority of the cases; exceptions do exist I'm sure).
 
Status
Not open for further replies.
Back
Top