Turtle Trading Rules

Chapter 10 Thinking Like a Turtle (2)
Think about it from a probabilistic perspective
Many people have taken probability and statistics courses in high school or college.

This graph reflects the distribution of female heights.The horizontal axis is the height value in inches, and the left and right vertical axes represent the following two probabilities:

1. Probability density: The shaded area is opposite to the scale on the left, which represents the probability corresponding to a specific height value.According to this chart, the average height of a woman is 64 inches (about 1.63 meters).The probability that a woman's height is close to this average is greater, and the further away from this average, the lower the probability.The highest point in the center of the graph is the greatest probability, and the lower areas on the sides represent smaller probabilities.For example, the height of the curve at 70 inches (about 1.78 meters) is much lower than that at 68 inches (about 1.73 meters), which means that a woman is much less likely to be 70 inches than 68 inches.

2.累积概率：图中的实线与右侧的刻度相对，从零一直延伸到100%。它表示一个女性的身高不高于某个水平的概率。比如，这条实线在70英寸左右的身高水平上升到了几乎100%的程度。确切数值是99.18%，这意味着只有不到1%的女性能达到70英寸以上的高度。

This graph and others like it are derived using complex mathematical formulas, but they all represent a simple concept: the farther you are from a central point that represents the mean, the lower the probability.

But why do we need to make the probability problem so complicated?Even if you ignore those mathematical knowledge and formulas, you can also use a simple method to make a picture like Figure 4-1: First, find a place with more women, such as a university campus; Next, randomly select Take 100 women, measure their heights; finally, divide these height data into several bins at intervals of 1 inch, and then count how many people are in each bin.If you do this, you are likely to find that there are 64 people at 16 inches, about 63 people each at 65 inches and 15 inches, about 62 people each at 66 inches and 12 inches, 61 people each at 67 inches and 8 inches, and 60 people at 68 inches. There are 8 people each in inches and 59 inches, 69 people each in 2 inches and 58 inches, and 70 person each in 1 inches and [-] inches.

If you made a bar graph based on the number of people at each particular height value, the graph would look like Figure 4-2:
It reflects the frequency of a particular metric value and other similar metric values ​​(in this case, female height).The shape of Figure 4-2 is similar to the normal distribution graph in Figure 4-1, but the advantage is that you can make such a graph without complex mathematical formulas, as long as you can count and classify.

This type of histogram can also be drawn from your trading system to give you a basic idea of ​​future trends; it can help you think about problems from a probability perspective instead of predicting something.Figure 4-3 is such a bar chart, which is the result of a 20-year monthly return test of a simplified version of the Turtle system, the Donchian Trend System.In addition to being simpler than the Turtle system, this system also outperforms the Turtle system.

Each bar represents the number of months corresponding to the corresponding grade: for example, the first bar on the right half indicates how many months the rate of return is between 2 and 2%, and the next bar represents the rate of return between 4% and [-] How many months are there between %, and so on.Note that the shape of this histogram is similar to the normal distribution of heights above.There is one notable difference though: the shape of this graph extends a longer distance to the right.This stretch represents better-performing months, and is also known in statistics as skew, or fat-tailed.

The left part is the losing trades and the right part is the winning trades.Note that each part has two vertical scales, one is the quantity scale at the left and right ends, and the other is the percentage scale in the middle, and the range is from 100 to 100%.The two cumulative distribution curves in the figure extend from the middle to the left and right sides respectively, rising from zero to [-]%.

On the quantity scale at the left and right ends, each bar represents 20% of losing or winning trades.For example, on the left end, the fifth box represents 3746 losing trades, or 100% of all losing trades.This means a total of 22 losing trades over the 3746-year test period.In the graph on the right, there are 1854 profitable transactions, accounting for 100% of the profitable transactions.

The bars in the figure represent the number of transactions with different R-multiples. The R multiplier is equal to the profit of a transaction divided by the risk investment of the transaction. This concept was invented by the trader Chuck Branscomb to compare the results of transactions between different systems and different markets. An easy way. (Van Tharp popularized the concept of the R multiplier in "Road to Freedom in the Financial Kingdom.") Here's an example.If you bought an August gold contract at $450 an ounce and your stop loss was at $8, your risk is $440 because the difference between $1000 and $450 ($440) is multiplied by one The amount of gold (10 ounces) represented by a contract is equal to $100.If the trade makes $1000, then it is a 5000R trade, where profitable trades are divided into different bins at 5R intervals, and losing trades are binned at 1/1R intervals.

It may seem strange that there are far more losing trades than winning trades in this graph, but it is actually a very normal phenomenon of trend following systems.However, although there are many losing trades, this system can ensure that the loss amount of most losing trades is close to the normal level, that is, the entry risk of 1R.In contrast, the profits of profitable trades are far greater than the risk of entering the market, and the return of 43 transactions is more than 10 times the risk of entering the market.

What does this tell us about our turtle minds?

The Turtles never knew if a trade would end up making or losing money.We only know the approximate shape of the distribution of trading outcomes - a shape very similar to the graphs above.We believe that every transaction has the potential to make money, but the most likely result is to lose money.We also know that some deals will have middling returns of 4R to 5R, and others will be big wins of 12R, 20R or even 30R.But in the end, it seems that the gains from the victories more than make up for the losses from the losses, and we always win.

Therefore, when we make a transaction, we don't judge our ability based on the result of this transaction, because we know that the greatest possibility is to lose money.We think in terms of probabilities, and because of that, we remain confident in the face of great risk and uncertainty.

think like a turtle

1. It's the present that matters: don't dwell on the past, and don't try to predict the future.The former is of no benefit to you, the latter is futile.

2. Think in terms of probability and don't predict the future: Don't try to make correct predictions, you can only succeed in the long run if you use the probability in your favor.

3. Take responsibility for your own trading results: Don’t blame other people, the market, your broker, etc. for your mistakes and failures.Take responsibility for your mistakes and learn from them.

Take responsibility for your own trading results
Some Turtles have a hard time accepting this concept, they want to be right and they want to be able to predict the future of the market.For this reason, even after the first month of heating oil trading provided a lesson, they still could not be steadfast in our system execution.I still remember that someone even suspected that Rich had taught me some laws alone that no other turtles knew.The idea is ridiculous.Why would Rich deliberately leave out some important information so the Turtles would lose his money?Not to mention that he would lose the bet with Bill.

There are no secret recipes.In fact, the method I actually use is much simpler than most other turtles.I put 100% of the money in my account into the 10-week breakout system with a longer time horizon.This means I have fewer trades and less monitoring of the market.I'm certainly not doing anything out of the ordinary, nor relying on any unpublished information.

For our skeptical Turtle, the easiest way to excuse our poor performance during the Turtle Project was to say that Rich had missed some important tip.This is a common problem in the trading world, and it's a common problem in life.Many people like to blame their failures on other people or on external circumstances beyond their control.They will blame everyone but themselves.This unwillingness to take responsibility for their actions and their consequences is perhaps their most important cause of failure.

The trading world is a great place to smash this bad habit.In the final analysis, trading is just a matter between you and the market, and you have nothing to hide from the market.If you do it well, you will see good results in the long run.If you do it poorly, you will lose money in the long run.But despite this obvious and inescapable connection between traders' behavior and results, some try to blame the market.Instead of taking responsibility for their own mistakes, they fantasized about spooky scenarios, suspecting that a group of "experts" or some other mysterious gang of traders had conspired to steal their money.

Of course, at any given time there will be plenty of traders trying to take your money, but I have never seen a collusion or fraud on this scale that these people conjured up out of thin air.Such people always blame the market, brokers, or other market participants for their mistakes.

You trade and you are responsible for the results.Don't blame anyone for giving you bad advice, and don't blame anyone for not telling you the secret.If you do something stupid, learn from your mistakes and don't act like you didn't make them.Then figure out how to avoid making the same mistake again.

(End of this chapter)