Trading Techniques   

The Trading Floor

If AI can learn on its own and teach itself and there are no parameters, the danger is what if AI starts to have its own value system
as to what is right and what is wrong OR what is good and what is bad OR who deserves to live and who doesnt... and it executes
based on its own judgment using its own value system. Then the human race would have created another race. 

bystander1965      ( Date: 19-Mar-2021 08:53) Posted: As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me. bystander1965      ( Date: 19-Mar-2021 08:48) Posted: I am talking about deep learning already. If you don' t feed it the N+1 dimension it will not look at it. When it can finally do inferences from completely unrelated areas very well, it will be a very scary day. As it stands, if you dig into some of the decision making process of the AI, even human cannot explain sometimes

 

Yup, dun forget a chimp throwing darts was a top fund manager... iirc ranked 22. Lol

bystander1965      ( Date: 19-Mar-2021 09:25) Posted: Anyway, it' s acknowledged that it' s rather difficult to use deep learning on stock markets due to its randomness. bystander1965      ( Date: 19-Mar-2021 09:21) Posted: Yes, I am just saying AI today, there are areas that it can' t do yet. You will be able to find similar things in the financial world as the car example. The algo they use do not just look at the chart/price data. A lot of them look at news also, what not. If it' s purely looking at the price data, then unless the data repeats itself consistently (and we kind of know it' s not like that exactly), the aglo may not be very efficient. Human will have to step in to tweak as the market condition changes

 

Anyway, it' s acknowledged that it' s rather difficult to use deep learning on stock markets due to its randomness.

bystander1965      ( Date: 19-Mar-2021 09:21) Posted: Yes, I am just saying AI today, there are areas that it can' t do yet. You will be able to find similar things in the financial world as the car example. The algo they use do not just look at the chart/price data. A lot of them look at news also, what not. If it' s purely looking at the price data, then unless the data repeats itself consistently (and we kind of know it' s not like that exactly), the aglo may not be very efficient. Human will have to step in to tweak as the market condition changes. BrainiakZ      ( Date: 19-Mar-2021 09:14) Posted: Just for discussion sake, not going against u or anything. Ur example requires hardware to get data, but in the financial world, data is readily available. So, it is not unforeseeable that there are systems out there that are trained to hunt for data that may fit a pattern. It may just tap on whatever is available on a bloomberg terminal at the simplest, or scour the entire search engine. I not an expert in the area, but as long as data is available i not surprised that this may be out there. Then again, those systems dun benefit us retailers so... we just hoot wat we can and huat as we hoot

 

Yes, I am just saying AI today, there are areas that it can' t do yet. You will be able to find similar things in the financial world as the car example.
The algo they use do not just look at the chart/price data. A lot of them look at news also, what not.
If it' s purely looking at the price data, then unless the data repeats itself consistently (and we kind of know it' s not like that exactly), the aglo may not be very efficient. Human will have to step in to tweak as the market condition changes.

BrainiakZ      ( Date: 19-Mar-2021 09:14) Posted: Just for discussion sake, not going against u or anything. Ur example requires hardware to get data, but in the financial world, data is readily available. So, it is not unforeseeable that there are systems out there that are trained to hunt for data that may fit a pattern. It may just tap on whatever is available on a bloomberg terminal at the simplest, or scour the entire search engine. I not an expert in the area, but as long as data is available i not surprised that this may be out there. Then again, those systems dun benefit us retailers so... we just hoot wat we can and huat as we hoot. bystander1965      ( Date: 19-Mar-2021 08:53) Posted: As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me

 

To add on, it also depends on the reits hedging and ability to pass on and increase rents. 

if looking at competing assets, it also depends what price was entered. If entered last mar, reits yield can tahan quite a bit.

bystander1965      ( Date: 19-Mar-2021 09:17) Posted: Rising rate is a risk for all businesses. And yes and no. Rate goes up because the market is anticipating a good economy ahead. Is that bad for businesses when the economy is good? As long as the rate doesn' t go up to such an extent that it hurts the businesses, it is a non issue. Before the crash in 2020, the yield was standing at 1.8% in early 2020. Before the crash in 2008, the yield was > 3.5%. It crashed to about 2.1% and bounced back to 3.8% by mid 2009. And we all know that' s the start of the 11-year bull run now.   halleluyah      ( Date: 19-Mar-2021 09:11) Posted: rising yield will be a risk to reits....

 

  Well as long as the AI Algorithms doesn' t " destroy" the hedge & pension fund houses, by plunging and short-selling every counter without remorse.....

bystander1965      ( Date: 19-Mar-2021 09:11) Posted: Cyborg days will be here. Haha. It' s like the holy grail for AI. We are looking for the holy grail in trading/investing. Still searching ... simpleguy123      ( Date: 19-Mar-2021 09:09) Posted: That' s what happened with SkyNet from the Terminator Franchise, it became " self-aware" that it was in a Master (Human) and Slave (AI) relationship. It didn' t help matters that SkyNet was programmed as a military weapon for tech cyber-warfare, where its main function was not really defense but offense against threats. And in this self-aware relationship, the Master (Human) was the threat to the Slave (AI)....

 

Rising rate is a risk for all businesses.
And yes and no.

Rate goes up because the market is anticipating a good economy ahead.
Is that bad for businesses when the economy is good?

As long as the rate doesn' t go up to such an extent that it hurts the businesses, it is a non issue.
Before the crash in 2020, the yield was standing at 1.8% in early 2020.
Before the crash in 2008, the yield was > 3.5%. It crashed to about 2.1% and bounced back to 3.8% by mid 2009. And we all know that' s the start of the 11-year bull run now.

 

halleluyah      ( Date: 19-Mar-2021 09:11) Posted: rising yield will be a risk to reits....

 

Most REITs are in red now.
A bit of panic?

 

Just for discussion sake, not going against u or anything. Ur example requires hardware to get data, but in the financial world, data is readily available. So, it is not unforeseeable that there are systems out there that are trained to hunt for data that may fit a pattern. It may just tap on whatever is available on a bloomberg terminal at the simplest, or scour the entire search engine. I not an expert in the area, but as long as data is available i not surprised that this may be out there. Then again, those systems dun benefit us retailers so... we just hoot wat we can and huat as we hoot.

bystander1965      ( Date: 19-Mar-2021 08:53) Posted: As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me. bystander1965      ( Date: 19-Mar-2021 08:48) Posted: I am talking about deep learning already. If you don' t feed it the N+1 dimension it will not look at it. When it can finally do inferences from completely unrelated areas very well, it will be a very scary day. As it stands, if you dig into some of the decision making process of the AI, even human cannot explain sometimes

 

Cyborg days will be here. Haha.
It' s like the holy grail for AI.
We are looking for the holy grail in trading/investing. Still searching ...

simpleguy123      ( Date: 19-Mar-2021 09:09) Posted: That' s what happened with SkyNet from the Terminator Franchise, it became " self-aware" that it was in a Master (Human) and Slave (AI) relationship. It didn' t help matters that SkyNet was programmed as a military weapon for tech cyber-warfare, where its main function was not really defense but offense against threats. And in this self-aware relationship, the Master (Human) was the threat to the Slave (AI).... bystander1965      ( Date: 19-Mar-2021 08:53) Posted: As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me

 

rising yield will be a risk to reits....

 

That' s what happened with SkyNet from the Terminator Franchise, it became " self-aware" that it was in a Master (Human) and Slave (AI) relationship.
It didn' t help matters that SkyNet was programmed as a military weapon for tech cyber-warfare, where its main function was not really defense but offense against threats. And in this self-aware relationship, the Master (Human) was the threat to the Slave (AI)....

bystander1965      ( Date: 19-Mar-2021 08:53) Posted: As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me. bystander1965      ( Date: 19-Mar-2021 08:48) Posted: I am talking about deep learning already. If you don' t feed it the N+1 dimension it will not look at it. When it can finally do inferences from completely unrelated areas very well, it will be a very scary day. As it stands, if you dig into some of the decision making process of the AI, even human cannot explain sometimes

 

MIT...

2.66 : 2.67 after touching 2.65.
Hmmm.... a chance to buy?   

 

See whether people buy the dips.
If dip buying is strong, she wont retrace much.
And if sell leow, may have to buy back higher.
If trading, I would probably jump IN and OUT and do HIT & RUN.
But I am doing BUY & HOLD.
So I need to do this.

.

spore1      ( Date: 19-Mar-2021 08:01) Posted: Hi, can take partial profit right!So, today likely go down with US over night lao Hong wavehunter      ( Date: 18-Mar-2021 23:44) Posted: Ya. Have been nibbling her. Hope she dont disappoint me. My prediction is she will show us 3.30 to 3.50 before the year is up. But what is unknown is how is she going to go there. Is she going to do the slow Tango from current level all the way till she reaches 3.30 (or higher) ..... OR... is she going to Lao Hong somewhere at 3.16 to 3.22 and go back down to the 2.80s-2.90s during a flushing operation to consolidate before heading back up again and this time towards my target zone of 3.30 to 3.50. That' s the one million dollar question. If dont sell at the Lau Hong and she goes back down, then wasted and NABEH. If sell and she no go down but continue to push higher towards my target zone but without me on board.... then also NABEH. Dilemma Dilemma. Shake head.

 

Isn' t it interesting that 10Y yield was increasing while DOW continues to go up? Maybe it' s artificial pushed to let go. We will never know TODAY.
Actually go find a chart where you can overlay the yield and DOW and see for yourself.

wavehunter      ( Date: 19-Mar-2021 08:53) Posted: Dow roae to an Intraday High of +212 last nite. But lost steam and closed down -153 pts. Small change. The bull is just crying " Woof !!!" Futures at +63. Observe the price action of the stocks you are looking to buy. Are people selling in panic? Does the selling trigger wave after wave of follow up selling after that? Or is the selling, if any, well absorbed becoz people are buying the dips? If your stock is a much sought after stock, you will see people buying the dips and selling is well absorbed. As a result, the price will hardly retrace and may even see a New High. If that is what you see, move in. If you dont want to miss your boat.  If cannot find the bola, then remain at the sidelines where you are safe.  wavehunter      ( Date: 18-Mar-2021 23:54) Posted: Dow is at +190 now. But the nite is still young though

 

As an example, let' s say for training a autonomous car, the driver each time ignores the speed bump and just drive fast over it. The AI will not learn to slow down when it comes to speed bumps. But for a human who has never seen a speed bump, the effect of running over it at speed is you get uncomfotable. The next time you will slow down. But the AI probably will never learn that unless it has gyrators inside the cars to say hey all my components are bouncing inside the car and they will wear down very quickly if you keep doing that. If the AI can automatically learn that that is bad, without human teaching it, wow, it will be scary to me.

bystander1965      ( Date: 19-Mar-2021 08:48) Posted: I am talking about deep learning already. If you don' t feed it the N+1 dimension it will not look at it. When it can finally do inferences from completely unrelated areas very well, it will be a very scary day. As it stands, if you dig into some of the decision making process of the AI, even human cannot explain sometimes. BrainiakZ      ( Date: 19-Mar-2021 08:42) Posted: There is also machine learning, so the algos may get smarter and learn faster than humans

 

Dow roae to an Intraday High of +212 last nite.
But lost steam and closed down -153 pts.
Small change. The bull is just crying " Woof !!!"
Futures at +63.

Observe the price action of the stocks you are looking to buy.
Are people selling in panic?
Does the selling trigger wave after wave of follow up selling after that?
Or is the selling, if any, well absorbed becoz people are buying the dips?
If your stock is a much sought after stock, you will see people buying the dips and selling is well absorbed.
As a result, the price will hardly retrace and may even see a New High.
If that is what you see, move in.
If you dont want to miss your boat. 
If cannot find the bola, then remain at the sidelines where you are safe. 

wavehunter      ( Date: 18-Mar-2021 23:54) Posted: Dow is at +190 now. But the nite is still young though. halleluyah      ( Date: 18-Mar-2021 11:26) Posted: most likely induce retailers to buy....bbs left hand to right hands to left hands....agak tonite US pull back, last nite up bcos of fed sweet tok if not mkt boh lat liao...

 

I am talking about deep learning already. If you don' t feed it the N+1 dimension it will not look at it. When it can finally do inferences from completely unrelated areas very well, it will be a very scary day.
As it stands, if you dig into some of the decision making process of the AI, even human cannot explain sometimes.

BrainiakZ      ( Date: 19-Mar-2021 08:42) Posted: There is also machine learning, so the algos may get smarter and learn faster than humans. bystander1965      ( Date: 19-Mar-2021 08:32) Posted: Nowadays it is not errors (unless it' s a bug). The machines are taught to look at N dimensions and decide. But there is a possibility that there is a N+1 dimension human has not taught them, and human only realises after seeing the situation. Human brains are still much more sophisticated but emotion gets in the way.

 

TikTalk      ( Date: 19-Mar-2021 08:10) Posted: VIX rose 2.35 to 21.58 as Nasdaq sell-off together with the broader S+P dragging down Dow. US Treasury yields continue to rise. Few days ago Bill Gross the founder of PIMCO also known as the Bond King said he was shorting US Treasury and that yield will rise above 2% this year. TikTalk      ( Date: 18-Mar-2021 10:00) Posted: NABEH & HENG RESEARCH - 18TH MARCH 2021 VIX : CBOE Market Volatility Index Pre Covid the lows of this " Fear" index was around 11, post Covid the lows for VIX has been around 20. Infact the lowest was 19.51 on Nov 27th. Which is no surprise you kept hearing bubbles......as VIX occassionally spike up from above 20 levels. Interesting......Last 2 sessions of VIX saw 2 new consecutive lows, 19.33 on Tuesday and 19.18 yesterday. IF this " fear index" continue to trend lower then you are likely to hear less of this word " bubbles" ...The Dow  will make new highs for each new lows of VIX. Keep an eye on VIX. Cheers Newbie TikTalk

 

There is also machine learning, so the algos may get smarter and learn faster than humans.

bystander1965      ( Date: 19-Mar-2021 08:32) Posted: Nowadays it is not errors (unless it' s a bug). The machines are taught to look at N dimensions and decide. But there is a possibility that there is a N+1 dimension human has not taught them, and human only realises after seeing the situation. Human brains are still much more sophisticated but emotion gets in the way.   kiseki_2818      ( Date: 19-Mar-2021 08:23) Posted: special situation human will need to quickly take control of the machine. bcos machine can do many things fast, esp when duplicationg errors. so machine and bbs can also lose $$$ to us