I did not intend to make this topic as a series despite I have made four posts about it already. However, since I have also written another related post regarding probability on trading and it is closely related to what I term it probability trading so it is good to supplement some more info about this idea in this post.
Every short side option trader knows that option trading boils down into three key estimations prior to position building,
1. Direction of index movement, ie., above or below the rollover point at the settlement day
2. Range of volatility and the peak and valley of the month
3. settlement point
Naturally there are also concern on position size, strike picking, time of position building and long or short which are more on the risk management and trade strategy which varies from one person to another.
In my first post I mentioned that as long as one believes the index movement is the manifestation of the combat between bearish and bullish master players so it follows a structured pattern which history will repeat in somewhat similarity. Therefore probability from the past statistic could be a good reference for the inference of the future. This is the basis of my idea of probability trading, ie., decision based on the likelihood on what things will/will not happen.
The application of probability on the above mentioned three key aspects does not share equal weighing. In fact probability shows little implication on the first aspect, ie., direction of movement, because whether the spot month index will end up higher or lower than the rollover point is indeed random and is affected by the key events happen in that month which probability data fails to predict. Having said so, statistic does reveal that for HSI when the previous month has experienced a more than 10% fall then there is a 0.78 probability that the spot month will see a higher than rollover point settlement.
Probability does it job best on the second aspect, ie., the prediction on the monthly volatility and peak/valley. Statistic shows that 0.87 probability monthly volatility is below 3000 points and 0.73 probability below 2000 points respectively in the past 18 years. Naturally how accurate the probability data for a particular month can be depends on the design of the database which I mentioned in my first post. A sophisticated database which incorporates well designed sorting parameters does reveal a reliable probability of the volatility of a month for a given pattern.
Although probability data cannot predict the exact settlement point for a particular month but just like what it does for the volatility, it can show the probability of the how many points deviation from the rollover point. For example, it is 0.05 probability that any given month the settlement is more than 1800 points above the rollover point.
In my post "is a trade worthwhile?", I raised the concept of odds and probability for the basis of whether a trade should be made at all. A well structured database can provide probability for the volatility and settlement point. The information of odds for index option trading for short side trader is even easier. Basically the profit for a short position is known when it is built. The potential loss depends on how the hedging is or what strike the protective long is put. The odds is just a simple maths.
As Charlie Munger puts it, value investing is all about looking for bets with 0.5 probability but with 3:1 odds. Through the use of database and the application of probability, index option trading can be more secured. The risk (probability of settlement beyond the strike of position) is foreseen and how much the profit is also known right at the time of position building. No more torture from the fear and greed. This is the essence of my idea of probability trading.
Good luck in trading!
2018年12月19日 星期三
2018年12月9日 星期日
Artificial Intelligence, a destructive goodie(2)
In my earlier post with the same subject I raised the concern on AI from the perspective of common folks. Not long ago I came across an article from a knowledgeable scholar on the topic on whether AI and robots can overtake human being. Contrary to my view, he is quite optimistic that AI will not surpass human despite of the former's ever expanding capability. His argument is that AI is silicon based while life on earth, in all forms including human's, is carbon based. While carbon based life has evolved into a stage that it has so many complexity of senses and feelings but AI even cannot perceive the feeling of hurt, so AI does not has the mental capacity like sensation, emotions and imagination that are needed to develop into the higher form of intelligence just like what human being has, let alone the development of self-consciousness that could turn out to be rivalry against human being.
In his view, even if silicon based AI can be evolved into life but it will be of no match with the carbon based life. His argument is that the existence of the millions species of carbon based life on earth now is the evidence that it is the best manifestation of natural selection should there ever had been any life-forming competition between the two chemical elements. That scholar is one of the respected persons in my life but I can't agree with him on this aspect.
Carbon based life is no doubt flourishing on this planet now but if we view the history on earth and also that of the universe, a million or even billion years is not more than just a blink of an eye. Dinosaur has once ruled the planet for 180 millions year but it is human being which is in control now. Considering the shear size of dinosaur, they would not think, if they ever did, this tiny primate can make such an achievement. Existing dominance might not and most probably cannot persist forever when the time line is long enough.
AI is still in its very infancy. Although we have seen much progress on its development but it is just like an embryo comparing to a fully grown adult if we put it to the metaphor of silicon based life to carbon based life. I am not sure whether AI will finally develop to has senses resemble to that of human's but I am pretty sure when AI's self learning runs long enough it will eventually has the awareness of self-consciousness. It is just like the process that a baby's brain will go through.
When AI is aware of self and external then it will unavoidably start thinking of the well-being of self. Silicon based life is then at the brink of emergence, cogito ergo sum (I think, therefore I am), if this philosophy applies. When there is life in a bacteria which does not think at all then we just cannot deny it is not life when silicon based AI can really think but not just playing algorithm. Naturally this form of life is quite different from that of the counterpart especially the way of reproduction.
I would say the relationship between the carbon-based life and the coming silicon based life is that the former develops a platform for the emergence of the latter. The essence of the existence of silicon based life is electricity. Without it no computer can ever run. Electricity was discovered not more than two centuries ago and it is human being ,ie carbon based life whom invented it. It is also human being who invented and nurtures AI to become a monster. Some may say AI is still under the control of human's hands but is it? Image one day Google, banking system, transport system, telecommunication system do not work simply we try to disconnect AI then what type of living will it become? Let alone AI at this stage is only in its primitive application. Can someone image how AI affects human's living 50 years or a century later? It is not difficult to think who is going to control whom.
The fact and reality are that it is a path of no return when the box of Pandora was opened. Human will rely on AI more than ever in the future. The point is that at certain point the race between carbon based life and silicon based life will become unequaled. The information processing capability of silicon based life is much much much much way stronger than its counterpart and unfortunately we are on the info/data highway which we build on our own. On the other hand, most of the human's knowledge and experiences vanish after death but AI is immortal and it learns and evolves in much faster rate than that of human's. Human replaced dinosaur to be the ruler of the world and then our civilization is replaced by that of the AI's whom dwarfs us because of the limitation on the design and structure of carbon based. It is just about evolution of history, in term of light year.
Our existence is for the preparation and to give rise to AI civilization. Human discovered electricity that flourishes us and it is just electricity undoing us.
In his view, even if silicon based AI can be evolved into life but it will be of no match with the carbon based life. His argument is that the existence of the millions species of carbon based life on earth now is the evidence that it is the best manifestation of natural selection should there ever had been any life-forming competition between the two chemical elements. That scholar is one of the respected persons in my life but I can't agree with him on this aspect.
Carbon based life is no doubt flourishing on this planet now but if we view the history on earth and also that of the universe, a million or even billion years is not more than just a blink of an eye. Dinosaur has once ruled the planet for 180 millions year but it is human being which is in control now. Considering the shear size of dinosaur, they would not think, if they ever did, this tiny primate can make such an achievement. Existing dominance might not and most probably cannot persist forever when the time line is long enough.
AI is still in its very infancy. Although we have seen much progress on its development but it is just like an embryo comparing to a fully grown adult if we put it to the metaphor of silicon based life to carbon based life. I am not sure whether AI will finally develop to has senses resemble to that of human's but I am pretty sure when AI's self learning runs long enough it will eventually has the awareness of self-consciousness. It is just like the process that a baby's brain will go through.
When AI is aware of self and external then it will unavoidably start thinking of the well-being of self. Silicon based life is then at the brink of emergence, cogito ergo sum (I think, therefore I am), if this philosophy applies. When there is life in a bacteria which does not think at all then we just cannot deny it is not life when silicon based AI can really think but not just playing algorithm. Naturally this form of life is quite different from that of the counterpart especially the way of reproduction.
I would say the relationship between the carbon-based life and the coming silicon based life is that the former develops a platform for the emergence of the latter. The essence of the existence of silicon based life is electricity. Without it no computer can ever run. Electricity was discovered not more than two centuries ago and it is human being ,ie carbon based life whom invented it. It is also human being who invented and nurtures AI to become a monster. Some may say AI is still under the control of human's hands but is it? Image one day Google, banking system, transport system, telecommunication system do not work simply we try to disconnect AI then what type of living will it become? Let alone AI at this stage is only in its primitive application. Can someone image how AI affects human's living 50 years or a century later? It is not difficult to think who is going to control whom.
The fact and reality are that it is a path of no return when the box of Pandora was opened. Human will rely on AI more than ever in the future. The point is that at certain point the race between carbon based life and silicon based life will become unequaled. The information processing capability of silicon based life is much much much much way stronger than its counterpart and unfortunately we are on the info/data highway which we build on our own. On the other hand, most of the human's knowledge and experiences vanish after death but AI is immortal and it learns and evolves in much faster rate than that of human's. Human replaced dinosaur to be the ruler of the world and then our civilization is replaced by that of the AI's whom dwarfs us because of the limitation on the design and structure of carbon based. It is just about evolution of history, in term of light year.
Our existence is for the preparation and to give rise to AI civilization. Human discovered electricity that flourishes us and it is just electricity undoing us.
2018年12月7日 星期五
Cornered
Soon after the long awaited meeting between the leaders of the U.S. and China after the G20 summit, there came the news that the CFO of Huawei, China's leading telecom equipment and services provider, was detained in Canada under the request of the U.S. and then many countries like Australia, Japan and New Zealand, following the U.S., have announced blocking on the purchase from Huawei. Apparently this is a concerted action among these countries and the purpose is also obvious. As mentioned in my first post about the trade war between China and the U.S. in this July that trade conflict is merely an excuse but the underlying issue lies on the combat on the global affairs leadership between the two nations.
This Huawei incident is for sure another battle in fighting for the say on the 5G telecommunication standard between the two countries so it is very clear why the U.S. prohibits the purchase on Huawei's equipments for any governmental agencies but why Australia, New Zealand and Japan follow suit? Indeed Europe, led by Germany, is also about to pull the welcome mat out from China's telecom companies. These countries are no match with China's strength on this aspect anyway so it appears to be weird. Some may suspect they are pressurized by the U.S. but except Japan, it is unlikely that these countries followed suit just because of the U.S.'s pressure. There must be some other reasons behind the concerted action.
The reason announced about the blockage is for the concern on national security so despite no explicit allegation on spying through the telecom equipment from China's companies but it was spoken between the lines. Apparently for outsiders it is only a Rashomon and China naturally denies. In fact unless either side can produce obvious evidence on their claims it will remain rashomon and the truth is not really the key issue. Anyone who has read some elementary psychology book know the concept of perception is truth. You are if I perceive you are no matter whether or not are you. This perspective holds true in international affairs as well. Despite of its open door policy 40 years ago till now, China remains suspicious to the West. After the collapse of the USSR she is the only communist country in the world that can post threat to the counterpart western world when Russia has already departed from the camp.
China's economic success made in the past 40 years is out of question a legend in human's history. The transformation from a laid back farming country to a manufacturing giant in such a short period has surprised the world. Her dark side is that she remains an authoritarian nation despite of her economic triumph. It is the main hindrance for China integrating into the world's main stream. Some Chinese die hard communism fans may still be suffered from the delusional disorder that the western world's wish to conquer us has never died so there is a resistant mindset against the modern mainstream values deep inside their subconsciousness. It is this kind of mentality that stops China having friend in the international community. Germany and Japan were once enemies to the U.S. but now they are allies. Many European countries had wars among each others for centuries but now they formed the EU. Why only China has got no true ally and is it the reason why she is cornered now? It is a billion dollars question.
Ideology is still a barrage and the East has never met the West.
This Huawei incident is for sure another battle in fighting for the say on the 5G telecommunication standard between the two countries so it is very clear why the U.S. prohibits the purchase on Huawei's equipments for any governmental agencies but why Australia, New Zealand and Japan follow suit? Indeed Europe, led by Germany, is also about to pull the welcome mat out from China's telecom companies. These countries are no match with China's strength on this aspect anyway so it appears to be weird. Some may suspect they are pressurized by the U.S. but except Japan, it is unlikely that these countries followed suit just because of the U.S.'s pressure. There must be some other reasons behind the concerted action.
The reason announced about the blockage is for the concern on national security so despite no explicit allegation on spying through the telecom equipment from China's companies but it was spoken between the lines. Apparently for outsiders it is only a Rashomon and China naturally denies. In fact unless either side can produce obvious evidence on their claims it will remain rashomon and the truth is not really the key issue. Anyone who has read some elementary psychology book know the concept of perception is truth. You are if I perceive you are no matter whether or not are you. This perspective holds true in international affairs as well. Despite of its open door policy 40 years ago till now, China remains suspicious to the West. After the collapse of the USSR she is the only communist country in the world that can post threat to the counterpart western world when Russia has already departed from the camp.
China's economic success made in the past 40 years is out of question a legend in human's history. The transformation from a laid back farming country to a manufacturing giant in such a short period has surprised the world. Her dark side is that she remains an authoritarian nation despite of her economic triumph. It is the main hindrance for China integrating into the world's main stream. Some Chinese die hard communism fans may still be suffered from the delusional disorder that the western world's wish to conquer us has never died so there is a resistant mindset against the modern mainstream values deep inside their subconsciousness. It is this kind of mentality that stops China having friend in the international community. Germany and Japan were once enemies to the U.S. but now they are allies. Many European countries had wars among each others for centuries but now they formed the EU. Why only China has got no true ally and is it the reason why she is cornered now? It is a billion dollars question.
Ideology is still a barrage and the East has never met the West.
2018年12月3日 星期一
Is a trade worthwhile?
A money gamer is constantly swinging between fear and greed. Greed is the primary drive for making a trade while fear always tortures the trader after making the trade especially when leverage is involved. As greed, ie., the desire to making profit, is the motivation of the trade while the fear comes from a loss resulting from that trade. After all, it is all about risk management.
There are many aspects in terms of risk management like the evaluation on macro economic situation eyes on the interest rate development, world economy and the fundamentals involving the outlook of the related asset or down to the trading strategy like the entry price, size of position, gearing and hedging etc. All these things are meant to deal with how a trade is executed. However, the most important point is still not yet addressed. It is whether or not the trade is worthwhile at all or should the trade be made at all.
There are some articles addressing how worthwhile a trade is by the concept of expected value (EV). Simply speaking, EV= (Probability of Gain) x (Take Profit Gain) ] + [ (Probability of Loss) x (Stop Loss Loss). According to those articles, higher the value better the trade is. However, it seems EV is good in comparing two or more alternatives of trade yet it still has not addressed to the point whether the trade is worthwhile or not unless the result is negative. To explain, below are two EVs with similar value but quite different scenario of winning/losing probability and the amount of gain/loss.
EV1=(0.1*$1000)+(0.9*-$10)=91.9
EV2=(0.9*$213)+(0.1*-$1000)=91.8
The two EVs are in similar value but do they give the punter similar risk and return? Apparently no! The scenario in EV1 resembles betting in lottery while EV2, as a local vivid metaphor puts it, win a candy but lose a factory (贏粒糖輸間廠). In EV1, the chance to win is slim (0.1 prob) but the reward is $1000 while a very high chance (0.9 prob) to lose but the loss is $10 only. The opposite is the EV2, despite of a high chance (0.9 prob) of winning however the reward is only $213 but as long as it loses then the loss is $1000 despite there is only 0.1 probability. In EV1, one risks his $10 to bet on a potential return of $1000 while in EV2 one is risking to lose $1000 in the hope to win $213 yet the EV value for the two cases are almost the same. Perhaps there is argument on the probability of the winning and losing in the two equations.
Probability is only the summary of the history of how likely the happening of something based on past record while history will not necessarily repeat in the future in the exact same fashion as it was in the past for a particular event. 0.1 probability of losing in EV2 means statistically there was one failure out from the 10 occasions in the past but it does not mean there must be one failure in every 10 occasions. For example 10 failures could scatter randomly throughout 100 occasions. There could be 2 failures in the first 10 occasions but none in the next 10 occasions yet the probability is still 0.1. If the failure just happened to fall onto the one the punter places his betting, so sad then. One must remember probability is good for the long run but absolutely not effective as what it suggests in any one particular occasion. Apart from this nature of probability but there is also black swan that makes probability is not so effective in term of reliability.
While it is good to win the jackpot in the lottery but we must avoid the scenario of winning a candy but to lose a factory when in failure. So when evaluating a trade(or a bet), the odds probably should be reviewed first. In EV1, the odds is 1000:10 while it is 213:1000 for EV2 or 100 vs 0.213 respectively. Judging from the odds alone then apparently EV1 is more favourable but there is only 10% chance to win the $1000 and 90% chance losing $10 in EV1 while it is 90% chance winning $213 and 10% chance losing $1000 in EV2. So what are the justified odds for these two sets of probabilities?
The formula to obtain odds from probability is O=1/(1-P) so the justified odds are 10 and 1.11 for EV1 and EV2 respectively. By comparing the offered odds of 100 and 0.213 for the two scenarios so it is clear that the trade(or bet) of wagering $10 in the hope of winning $1000 with just 0.1 probability is still a good deal and thus worthwhile. Simply speaking, only when the offered odds is greater than that derived from the winning probability then the trade (or bet) is worthwhile. In fact this is the underlying principle of value investing as remarked by Charlie Munger that value investing is looking for a bet with 0.5 winning probability but odds at 3:1 and hoping to find this misquoted odds. I think many of the so-called value investors do not even know this principle so resorting in nagging at the intrinsic value of a stock.
By considering the odds and probability altogether perhaps can give a better evaluation on how worthwhile a trade(or bet) is than can the Expected Value.
There are many aspects in terms of risk management like the evaluation on macro economic situation eyes on the interest rate development, world economy and the fundamentals involving the outlook of the related asset or down to the trading strategy like the entry price, size of position, gearing and hedging etc. All these things are meant to deal with how a trade is executed. However, the most important point is still not yet addressed. It is whether or not the trade is worthwhile at all or should the trade be made at all.
There are some articles addressing how worthwhile a trade is by the concept of expected value (EV). Simply speaking, EV= (Probability of Gain) x (Take Profit Gain) ] + [ (Probability of Loss) x (Stop Loss Loss). According to those articles, higher the value better the trade is. However, it seems EV is good in comparing two or more alternatives of trade yet it still has not addressed to the point whether the trade is worthwhile or not unless the result is negative. To explain, below are two EVs with similar value but quite different scenario of winning/losing probability and the amount of gain/loss.
EV1=(0.1*$1000)+(0.9*-$10)=91.9
EV2=(0.9*$213)+(0.1*-$1000)=91.8
The two EVs are in similar value but do they give the punter similar risk and return? Apparently no! The scenario in EV1 resembles betting in lottery while EV2, as a local vivid metaphor puts it, win a candy but lose a factory (贏粒糖輸間廠). In EV1, the chance to win is slim (0.1 prob) but the reward is $1000 while a very high chance (0.9 prob) to lose but the loss is $10 only. The opposite is the EV2, despite of a high chance (0.9 prob) of winning however the reward is only $213 but as long as it loses then the loss is $1000 despite there is only 0.1 probability. In EV1, one risks his $10 to bet on a potential return of $1000 while in EV2 one is risking to lose $1000 in the hope to win $213 yet the EV value for the two cases are almost the same. Perhaps there is argument on the probability of the winning and losing in the two equations.
Probability is only the summary of the history of how likely the happening of something based on past record while history will not necessarily repeat in the future in the exact same fashion as it was in the past for a particular event. 0.1 probability of losing in EV2 means statistically there was one failure out from the 10 occasions in the past but it does not mean there must be one failure in every 10 occasions. For example 10 failures could scatter randomly throughout 100 occasions. There could be 2 failures in the first 10 occasions but none in the next 10 occasions yet the probability is still 0.1. If the failure just happened to fall onto the one the punter places his betting, so sad then. One must remember probability is good for the long run but absolutely not effective as what it suggests in any one particular occasion. Apart from this nature of probability but there is also black swan that makes probability is not so effective in term of reliability.
While it is good to win the jackpot in the lottery but we must avoid the scenario of winning a candy but to lose a factory when in failure. So when evaluating a trade(or a bet), the odds probably should be reviewed first. In EV1, the odds is 1000:10 while it is 213:1000 for EV2 or 100 vs 0.213 respectively. Judging from the odds alone then apparently EV1 is more favourable but there is only 10% chance to win the $1000 and 90% chance losing $10 in EV1 while it is 90% chance winning $213 and 10% chance losing $1000 in EV2. So what are the justified odds for these two sets of probabilities?
The formula to obtain odds from probability is O=1/(1-P) so the justified odds are 10 and 1.11 for EV1 and EV2 respectively. By comparing the offered odds of 100 and 0.213 for the two scenarios so it is clear that the trade(or bet) of wagering $10 in the hope of winning $1000 with just 0.1 probability is still a good deal and thus worthwhile. Simply speaking, only when the offered odds is greater than that derived from the winning probability then the trade (or bet) is worthwhile. In fact this is the underlying principle of value investing as remarked by Charlie Munger that value investing is looking for a bet with 0.5 winning probability but odds at 3:1 and hoping to find this misquoted odds. I think many of the so-called value investors do not even know this principle so resorting in nagging at the intrinsic value of a stock.
By considering the odds and probability altogether perhaps can give a better evaluation on how worthwhile a trade(or bet) is than can the Expected Value.
訂閱：
文章 (Atom)