Randomness of price values - page 7

 
hrenfx:

Talked to US hedge fund strategists...

The server racks are at the exchange "behind the wall". Completely own software, which is written in C++ under CentOS. Speed is everything.

Own strategy tester not just on tick data history, but on the history of changes in Level2-stock on thousands of financial instruments.

The entire NYSE+NASDAQ is traded. Portfolio > $1 bn. Market-neutral where possible, i.e. no difference in how the market behaves.

Commission is negative. I.e. if they limit opened and limit closed at the same price, they are on the plus side - the exchange pays them extra. Because they create more liquidity.

1. All strategies (many) are intraday. Positions are not transferred to the next day. 1.The strategies are fully automatic.

No analysis of price BPs quantified by time. For example, timeframe BPs. No indicators.

No financially educated people among the developers. Exceptionally excellent mathematicians (PhD and postdoctoral) with excellent programming experience.

Complete conviction that the price behaves randomly. And there is no way to predict it for more than a day. Other examples are considered as luck and chance.

Working without the use of insider knowledge. But 2. unspoken features of pricing on exchanges are used.

Daily completely closed in-house discussions of new ideas for strategies.

Sometimes sincere regrets about 3. the delusions of traders on the forums and their complete lack of understanding of what they are dealing with .


1. Generally speaking, what are the strategies. BPs do not analyse, do not use indicators, work intraday. What are they working on?

2. is this an answer to p.1?

3. Delusions about what? Opinions abound on the forums...

 
sever30:
Said enough to create a performance.
 
sever30:


1. In general, we would like to know what strategies. They do not analyse BP, they do not use indicators, they work intraday. What do they work on?

2. is this the answer to point 1?

3. Delusions about what? Opinions abound on the forums...

Have you tried pipsing ( on funds ) ? Try it, i think it will take you about a week to start trading in profit.... i learnt on day 4 and still trading. This way the questions will fall away and you yourself will answer the question about the need for indicators. In a few seconds (or half a second) you see the beginning of growth, even before the price moves - and that's all you need to know to make money.

You will not make billions, but it is real money to live on and there is no magic. There is nothing complicated about it.

You don't need any fucking professors for that, and you don't need an excellent knowledge of mathematics and programming. Vat nothing is needed (just if you make it more convenient to trade with their software). There is nothing complicated about it. All they say "and they say that ..." but you try it, after 3 months you buy yourself a new car and forget about BP and the Efficient Market.

 

hrenfx, about those three slides - at least give me a link to the presentation and video for the sake of decorum (for the rest of the forum visitors; I know the source)

 
Presentation. Video presentation.
 
hrenfx:

Talked to US hedge fund strategists...

The server racks are at the exchange "behind the wall". Completely own software, which is written in C++ under CentOS. Speed is everything.

Own strategy tester not just on tick data history, but on the history of Level2-stock changes on thousands of financial instruments.

The entire NYSE+NASDAQ is traded. Portfolio > $1 bn. Market-neutral where possible, i.e. no difference in how the market behaves.

Commission is negative. I.e. if they limit opened and limit closed at the same price, they are on the plus side - the exchange pays them extra. Because they create more liquidity.

All strategies are (a lot) intraday. Positions are not transferred to the next day. Strategies are fully automatic. Some strategies are based on analysis of changes in the bids of the nearest price levels. Most of the strategies belong to high-frequency type - High-Frequency Trading.

No analysis of quantized in time price BPs. For example, timeframe BPs. No indicators.

No financial people among developers. Exceptionally excellent mathematicians (Ph.D. and postdoctoral) with perfect programming experience.

Complete conviction that the price behaves randomly. And there is no way to predict it for more than a day. Other examples are considered as luck and chance.

An operation without the use of insiders. But unspoken features of price formation on exchanges are used.

Daily completely closed in-house discussions of new strategy ideas.

Sometimes sincere regrets about traders' misconceptions on the forums, and their complete lack of understanding of what they are dealing with.

There is a perception that strategies should not contain very complicated mathematics. Maximum complexity at the level of covariance matrices. Sceptical, as practitioners, of the results of probability theory and those close to it.

Adequately and with interest to fresh and interesting in their opinion ideas. Willing to cooperate.


Well, it's their job to do it. They have an advantage for the HFT, so they implement it. But that doesn't mean they are right about it: "Full conviction that the price behaves randomly. And there is no way to predict it for more than a day. Other examples are seen as luck and chance". Interview successful mutual fund managers and they will tell you that price is completely random within a day. And what, you have to believe them all? :)
 
hrenfx:

Talked to US hedge fund strategists...

The server racks are at the exchange "behind the wall". Completely own software, which is written in C++ under CentOS. Speed is everything.

Own strategy tester not just on tick data history, but on the history of changes in Level2-stock on thousands of financial instruments.

The entire NYSE+NASDAQ is traded. Portfolio > $1 bn. Market-neutral where possible, i.e. no difference in how the market beh aves.

Commission is negative. I.e. if they limit opened and limit closed at the same price, they are on the plus side - the exchange pays them extra. Because they create more liquidity.

All strategies are (a lot) intraday. Positions are not transferred to the next day. Strategies are fully automatic. Some strategies are based on analysis of changes in the bids of the nearest price levels. Most of the strategies belong to high-frequency type - High-Frequency Trading.

No analysis of quantized in time price BPs. For example, timeframe BPs. No indicators.

No financial people among developers. Exceptionally excellent mathematicians (Ph.D. and postdoctoral) with perfect programming experience.

Complete conviction that the price behaves randomly. And there is no way to predict it for more than a day. Other examples are considered as luck and chance.

An operation without the use of insiders. But unspoken features of price formation on exchanges are used.

Daily completely closed in-house discussions of new strategy ideas.

Sometimes sincere regrets about traders' delusions on the forums, and their complete lack of understanding of what they are dealing with.

There is a perception that strategies should not contain very complicated mathematics. Maximum complexity at the level of covariance matrices. Sceptical, as practitioners, of the results of probability theory and those close to it.

Adequately and with interest to fresh and interesting in their opinion ideas. Willing to cooperate.

This boldface contradicts the claim that the market is random.

On the contrary, it proves that the regularities in the market are actively used, and the analysis of as many instruments as possible is used for this purpose. And the faster the pattern is detected, the more profit can be made. This is what superpowered computers are for. One does not need a computer at all to open in a random direction - a coin polished on both sides will do.

In fact, what kind of an insider can be if the market is random? :)

 
hrenfx:

Talked to US hedge fund strategists...


I'd like some confirmation of what you're saying.

Communicated yourself or is this a reprint from somewhere? Can I get a link? If you did, how did you get involved with them?

How did you determine that they were "American hedge fund strategists"?

 
hrenfx:

Talked to US hedge fund strategists...

The server racks are at the exchange "behind the wall". Completely own software, which is written in C++ under CentOS. Speed is everything.

Own strategy tester not just on tick data history, but on the history of Level2-stock changes on thousands of financial instruments.

The entire NYSE+NASDAQ is traded. Portfolio > $1 bn. Market-neutral where possible, i.e. no difference in how the market behaves.

Commission is negative. I.e. if they limit opened and limit closed at the same price, they are on the plus side - the exchange pays them extra. Because they create more liquidity.

All strategies are (a lot) intraday. Positions are not transferred to the next day. Strategies are fully automatic. Some strategies are based on analysis of changes in the bids of the nearest price levels. Most of the strategies belong to high-frequency type - High-Frequency Trading.

No analysis of quantized in time price BPs. For example, timeframe BPs. No indicators.

No financial people among developers. Exceptionally excellent mathematicians (Ph.D. and postdoctoral) with perfect programming experience.

Complete conviction that the price behaves randomly. And there is no way to predict it for more than a day. Other examples are considered as luck and chance.

An operation without the use of insiders. But unspoken features of price formation on exchanges are used.

Daily completely closed in-house discussions of new strategy ideas.

Sometimes sincere regrets about traders' misconceptions on the forums, and their complete lack of understanding of what they are dealing with.

There is a perception that strategies should not contain very complicated mathematics. Maximum complexity at the level of covariance matrices. Sceptical, as practitioners, of the results of probability theory and those close to it.

Adequately and with interest to fresh and interesting in their opinion ideas. Willing to cooperate.

 
hrenfx:

The commission is negative. That is, if they limit open and limit close at the same price, they are on the plus side - the exchange pays them extra.

...

No analysis of quantified timeframed price BPs. Such as timeframe BPs. No indicators.


There are no people with a financial background among the developers.

Do they need it? Under such conditions with positive expected payoff, they do not have to do anything, except artificially maintain liquidity. They are not traders and investors and even not market makers, but exchange specialists. There is no risk in their activity, because they only get paid for their work.

Professionals don't give a shit whether price behaves randomly or not.