Reading notes
Title: The Fool Who Walks Randomly
Chapter 2, Chapter 4
Chapter Name: We are all probability blind
Author: Nassim. Nicholas. Taleb (Masterpiece: Black Swan)
We naturally like certain things and reject uncertain things, so we are not born with enough understanding of randomness, so it is easy to be fooled by randomness. Probability is an abstract concept, and human intuition is difficult to understand. In this chapter, Taleb gives many examples, but in fact, he also says several solutions:
1, the use of people's psychological characteristics for positive or reverse operation
2, understand the real probability in order to overcome the fear in the heart, and even profit from it, such as options
3. Stay away from noise and avoid being misled by wrong information
4, master the scientific knowledge of probability, in order to reduce the chance of being deceived.
Examples in life use this property for momentum or reverse manipulation
Taleb gives an example, for example, if you're going on vacation, maybe Paris, maybe the Caribbean, you might have a picture of two places in your head, but the probability of each being 50 percent obviously doesn't come to your head. Or what about 85% and 15%?
Another example is the concept of expectation value, 1000 participation in gambling, 50% probability to get zero yuan, 50% probability of 2000 yuan, then the expected value = 50% * 0 + 50% * 2000 = 1000.
For example, in Talipo's cancer, doctors say that the 5-year survival rate is 72%, which actually means that there is a 72% chance of being alive and 28% of the time.
There are two types of funds: one is the contrarians, whose logic is: Hey, because people will systematically overreact, so we come and fight them, sell the winners, buy the losers. The other is momentum players, who behave in the opposite way, because the market doesn't adjust fast enough, so we buy the winners and sell the losers. Because of randomness, both kinds of people can gain something in certain periods, but it is not possible to judge who is right and who is wrong in the two theories.
A legal example, misestimating the likelihood of a crime
Legal examples: probability of evidence of crime, probability of crime, Simpson. A Harvard-based lawyer made a specious argument that only 10 percent of men who abuse their wives kill them. This is a probability that is not conditioned (premised) on murder. The correct approach should be to look at the rate (up to 50 per cent) of wives who have been murdered by their husbands in the past. Because we are talking about conditional probabilities, that is, the probability that Simpson committed murder on the condition that Simpson's wife was killed, not the unconditional probability of Simpson's wife killing. It shows that university professors will understand the probability of error, not to mention ordinary people.
There is also the concept of joint probability, I was diagnosed with respiratory cancer, and the probability of being killed by a pink Cadillac sedan in the same year, assuming that it is 1 in 100,000 each, the probability that the two will hold at the same time is 10 billionths, that is, multiplying the probabilities of two apparently independent events. From the point of view of blood type, simpson has a 500,000 chance of not being a murderer (the lawyer uses this sophistry to say that there are 4 people of this blood type in Los Angeles), plus he is the husband of the victim, and other evidence, then due to the multiplier effect, the probability of being unfavorable to him will be much higher.
Kafka's novel The Trial describes some of the absurdity of a man who lacks the concept of probability, and then Taleb hacks lawyers and economists, saying that lawyers don't understand nonsense, that economists use mathematics in the wrong places, and that they take their theories too seriously.
Examples in psychology, misestimating the probability of illness, the risk of options
Psychologists Kahneman (author of Thinking Fast and Slow) and Watsky (co-authored The Discovery of Thinking with Kahneman) teach us how to feel and deal with uncertainty. They found that some of the physiology of the human brain leads us to perceive things and exhibit certain behaviors in certain established ways. The reason is that in the process of human evolution, because we don't have to understand probabilities, there are not enough incentives to develop the ability to understand probabilities. But the deeper reason is that humans are not designed to understand things, we are only designed to survive and reproduce, but in order to survive, we must exaggerate the probabilities of certain things, such as the probability of events that may affect our survival. Therefore, people whose brains are particularly concerned about the danger of life (that is, paranoia) are more able to survive. (Isn't this crisis awareness, bottom-line thinking?) ）
Taleb then introduces Bennett's Are You Right? In one example, testing for a disease has a 5% chance of producing false positives. The entire population has a 1‰ chance of developing the disease. Whether or not the disease is suspected, after randomly testing a group of people, it is found that one patient has a positive test result, so what is the probability that the patient will contract the disease?
Assume that no false negatives exist. One of the 1,000 patients tested is expected to develop the disease. Of the 999 healthy patients, about 50 will be tested because the accuracy rate is 95%. The correct answer should be that someone randomly selected has the following probability of testing positive and indeed contracting the disease: the answer is 1 in 51. (This result is indeed beyond imagination)
Taleb popularized his option winning method again: Suppose a stock is trading at $100 today, and someone gives me the right (but not the obligation) to buy it for $110 a month later. This is called a call option. A month later, the stock price rose above $110, and I asked the option seller to sell me the stock for $110, called exercise. That's when it makes sense to fulfill the contract. If the stock price at the time of the performance is $120, then my option has a value of $10, because I can buy the stock from the option seller at $110 and then sell it on the market at $120, with a difference of $10 in and a gap of $10. However, the probability of this is not very high. When buying a call option, if the market price of the stock is lower than the strike price, there is no benefit to immediately performing the option, which we call out-of-the-money.
Suppose I buy this buy option at $1. So how much is expected to be the value of this option a month from now? Most people think it's zero. This answer is not true. There is a high probability (say, 90%) that the value of this option expires to zero, but there may be a 10% probability that the average value is $10. So the seller sells my options and charges me $1 for free. If the seller does not sell my options, but buys the stock at $100, waits for 1 month, and can sell it for $120 and make a profit of $20. So making $1 now is really nothing. Similarly, the options I bought are not depletive assets, but even professionals can be fooled. Why? Because they confuse the expected value with the most likely scenario, in this case, the option has an expected value of $1, and the most likely scenario is a value of zero. Psychologically, they overestimate the importance of the most likely scenario, that is, the belief that the market will not move. The value of this option is actually a weighted average of each possible condition.
Examples from the press, stay away from noise
There are also some examples of the press, I simply listed down, do not expand, interested can see the original text:
Taking Fluorouracil to treat upper respiratory cancer found that it was better than placebo, but the difference was not great; with other physical therapies, it improved survival from 21% to 24%. Since my sample was small, the survival rate increased by 3%, and I'm not sure if it was due to drugs, as it could be just random.
The broader market only fell 10% from its highs, but individual stocks fell by an average of nearly 40% from their highs.
The average American is expected to live to be 73 years old. So if you're 68 and you can live for 5 years, you should plan your investment for the next 5 years.
The Dow Rose 1.03 points as interest rates fell
The dollar fell $0.12 as Japan's trade surplus widened
Causality can be very complex. If there are many possible causes, you cannot pick one alone as the cause. This is called multivariate analysis. For example, if the stock market is likely to be affected by domestic interest rates, the dollar/yen exchange rate, the dollar/euro exchange rate, the European stock market, the Balance of Payments account in the United States, inflation in the United States, and a dozen other major factors, then journalists must observe all of these factors, observe their previous individual and collective effects, observe the stability of these effects, and conduct statistical tests to show whether the rise and fall of the stock market is caused by a certain factor. Finally, the factors must be given a confidence level; if the level of trust is less than 90%, there is nothing to say.
In the opinions expressed by a person, the estimated or predicted value is not as important as the level of trust. Let's say you're on a fall morning and you have to know what the weather is like before you pack your bags. If you expect the temperature to be 18 degrees Celsius, plus or minus 5 degrees Celsius (for example, in Arizona), then you won't bring heavy snowsuits and portable fans. If you're going to Chicago and someone tells you that the temperature there, although 18 degrees Celsius, may vary to 20 degrees Celsius, then what do you do? You have to bring both winter and summer clothes. When preparing clothes, it does not matter what the temperature expectation is, what is important is the high and low temperature difference. When you know that the difference between high and low temperatures is 40 degrees Celsius, the clothes you prepare are very different. Further, if you are going to a planet, the expected temperature is also about 18 degrees Celsius, but add or subtract 200 degrees Celsius, how do you prepare clothes at this time? So, when operating in the market, the importance of where the market will go is far less important than the error rate I allow at a certain level of trust.