Turtle Trading Rules

Chapter 11 Discovering System Strengths
Having an edge is the difference between an expert and an amateur.Ignore this and you will lose out to those who have not.

The so-called successful trading is nothing more than buying at a low point and then selling at a high point, or selling short at a high point and then buying back at a low point to close the position.When deciding when to enter the market, most of the methods used by novices are similar to luck.In the eyes of experienced traders, there is no advantage to their approach.The word edge is borrowed from gambling theory, originally referring to the statistical advantage held by the casino, and it also refers to the advantage that a blackjack player may gain by counting cards.In gambling, if you don't have an edge, you're sure to lose in the long run.

The same is true in the trading world.If you don't have an edge, the cost of trading can cost you money.Commissions, transaction price deviations, computer costs, transaction fees and price data fees can skyrocket.An edge in trading is an exploitable statistical advantage based on market behavior that is recurring.In the trading world, the best edge comes from market behavior driven by human cognitive biases.

Three Elements of System Advantage

To find an edge, you need to find an entry point where there is a greater than normal probability that the market will move in a particular direction within an ideal time frame.Then, you have to design an exit strategy for this entry point so that you can profit from the desired trend.Simply put, entry and exit strategies must be paired for maximum advantage.Thus, a trend following entry strategy can be paired with many different types of trend following exit strategies, a counter trend entry strategy can be paired with many different types of counter trend exit strategies, and a swing trading entry strategy can be paired with many different types of swing trading exits. Strategy pairing, and more.

To understand the importance of this, let's look at what makes up a system's strengths.System advantage comes from three elements:

Portfolio selection: the algorithm for deciding which markets to enter.

Market Entry Signal: An algorithm that determines when to open a trade.

Exit Signal: An algorithm that determines when to exit a trade.

It is possible for an entry signal to have significant advantage in the short term, but not in the medium or long term.And vice versa, an exit signal may be advantageous for a long-term system but not for a short-term system.A few examples help us understand this effect.

odds ratio
When a specific market behavior occurs, the system will send a signal to enter the market.When you examine entry signals, you need to focus on the price movement that accompanies this market action.We can divide this price movement into two parts: good movement and bad movement.

A good change is one that works in your favor.In other words, if you buy, the market goes up is a good move, and the market goes down is a bad move; and if you sell short, a down market is a good move, and a go up is a bad move.To give an example: after buying, the market first moves in a bad direction, that is, it falls; then, the market rebounds and rises to a new height, which is higher than the initial purchase price; then, the market turns again , began to fall, as shown in Figure 5–1.

Traders call the maximum excursion in the bad direction MAE (maximum adverse excursion, maximum adverse excursion), and the maximum excursion in the good direction as MFE (maximum favorable excursion, maximum favorable excursion).Therefore, the double-arrowed lines in the figure indicate the MAE and MFE of this price movement.In this example, MFE (good price movement) is much larger than MAE (bad price movement).

You can use these concepts to directly measure the strength of an entry signal.Looking at the price movement after an entry signal, if the average maximum change in the positive direction is greater than the average maximum change in the bad direction (that is, the average MFE is higher than the average MAE), this indicates that there is a positive advantage.If the average MAE (Unfavorable Movement) is higher than the average MFE (Favourable Movement), this indicates a negative edge.It can be said that a truly random entry strategy will produce roughly equal amounts of good and bad moves.For example, if you make buying and selling decisions by tossing a coin, buy if heads come up, and sell when tails come up, then MFE and MAE should be equal in price changes after you enter the market.

This is just one way of thinking about entry advantage, and in order to turn it into an actual way of measuring entry signal advantage, we have to add a few more steps.First, you have to find a way to make price movements in different markets comparable.Second, you have to figure out which period to measure the average MFE and MAE.Comparing these averages is meaningful only if the MFE and MAE of different markets are normalized.To do this, you can borrow what the Turtles used to normalize trade sizes across markets: unify them with the Average True Range (ATR).

In order to clearly distinguish the entry performance in different markets, it is also necessary to compare the effect of an entry signal in different time periods.I usually test for a specific number of days, so I measure the MFE and MAE for a specific number of days after each entry signal.I am currently the Head of Research and Development at Trading Blox, a firm with a state-of-the-art system testing environment, where we have invented a measure of market entry advantage that we call the E-ratio (short for odds ratio).The E-ratio combines all of the above elements through the following formula:

1. Calculate the MFE and MAE for each market entry signal for the specified time period.

2. Divide each of the above MFE and MAE values ​​by the ATR at the time of market entry. This is to adjust for volatility and standardize different markets.

3. Sum the above adjusted MFE and MAE values ​​respectively, and then divide by the total number of market entry signals to obtain the adjusted average MFE and MAE.

4. The adjusted average MFE divided by the adjusted average MAE is the E-ratio.

In order to illustrate the time problem, we will indicate the number of days in the expression of E-ratio, that is, the time period for calculating MFE and MAE.For example, the E10-ratio uses MFE and MAE for 10 days, including the day of entry; the E50-ratio uses MFE and MAE for 50 days, and so on.

The E-ratio can be used to measure whether an entry signal has an advantage.For example, you can use it to test whether a completely random entry strategy has an advantage.Here's an example: I examined the E-ratio over the past 10 years with a randomized entry strategy that determines whether to open long or short based on a random computer process (the equivalent of a coin toss). The average result of 30 tests shows that the E5-ratio is 1.01, the E1-ratio is 1.005, and the E50-ratio is 0.997.As we expected, these numbers are very close to 1.0, and as we increase the number of experiments, these numbers get closer and closer to 1.0.The reason is simple: for a position, within any reasonable time frame, the price may change in a good direction or in a bad direction, and the probability of the two is [-]-[-]%.

You can also use the E-Ratio to examine the main elements of the Donchian Trend System.The market entry strategy of this system has two major elements, one is the Donchian channel breakthrough rule, and the other is the trend combination filter.According to the Donchian channel breakout rule, you should buy when the price breaks the highest point in the past 20 days, and sell short when the price breaks the lowest point in the past 20 days.The trend combination filter stipulates that you can only go long in markets where the 50-day moving average is above the 300-day moving average, and go short in markets where the 50-day moving average is below the 300-day moving average.If the state of a market is unfavorable to the system, the trend combination filter will remove the market.

让我告诉你如何用E–比率来检验唐奇安趋势系统的入市法则。下面所说的所有检验结果都是以28个高容量的美国期货市场为样本，采用的是1996年1月1日至2006年6月30日的数据。

The E5-ratio for our sample is 0.99 and the E10-ratio is 1.0. "Wait," you might say, "if an entry strategy has a positive edge, then the E-ratio should be greater than 1." Indeed it is.However, please don't forget that the Donchian channel breakthrough system is a medium-term trend-following system, so its market entry strategy can only show advantages in the medium term, not the short term.More generally, the advantages of an entry strategy are only manifested within the time horizon to which the strategy corresponds.

我们这个入市策略的E70–比率是1.20。这意味着，对20日突破法交易来说，入市信号之后70天内的平均有利变动幅度比平均不利变动幅度大20%。

It reflects the advantage ratio of the 20-day breakthrough system in different time spans.Initially, the odds ratio is less than 1.0, which indicates that unfavorable moves in short periods of time are generally larger than favorable moves.This is one of the reasons why it is so difficult for breakout traders to overcome psychological barriers.It is also for this reason that you can make money using counter-trend trading strategies, hoping that a breakout trend will not hold and instead bounce off a support or resistance mechanism.

Next, the odds ratio began to climb steadily, but still fluctuated irregularly in the region above 1.0.This suggests a positive advantage, but it is difficult to quantify it precisely.

Trend Combination Filter

How do portfolio selection criteria affect the strengths of the Donchian channel system?You can check this in two ways.First, you can look at what the trend combination filter does to the dominance of a purely stochastic entry strategy, and compare that result to 1.0 - because without any filtering, the stochastic strategy's base dodge ratio should be is 1.0.Second, you can combine the filter with our entry signal to see what effect the trend combination filter has on the odds ratio of our breakout signal.

Examining 7 random entry signals combined with asset filters, we obtain an E1.27–ratio of 70, higher than the E70–ratio of the entry signals themselves.This clearly shows that this portfolio selection rule can enhance the advantages of the system.

Using the Trend Combination filter greatly increases the likelihood of desirable results in breakout trades.In our example, the E70–ratio changed from 1.20 to 1.33.Also, if you combine the trend combination filter with the breakout method, the shape of the odds ratio graph changes and becomes smoother.

After we added the Trend Combination filter, the graph got a lot taller and smoother.As can be seen in the figure, the E120– ratio reached about 1.6.

This is so because breakout trades that go against the long-term trend are all removed.Many significant moves against initial positions are made by these trades, as breakouts against a major trend are difficult to sustain for extended periods of time.The occurrence of these breakouts also shows that the market is in a state that is not conducive to Donchian's trend system.

Advantages of an exit strategy
Even a systematic exit strategy should have advantages, if at all possible.Unfortunately, measuring the merits of an exit strategy is even more difficult.This is because the exit strategy is related to both the entry strategy and the exit signal.In other words, you cannot examine exit strategies in isolation from the reasons for entering a position in the first place.There is not just one system element, but many different elements with intricate interactions between them.

Odds ratio combined with trend filter

Since this system is very complex, instead of caring about the advantages of the exit strategy, it is better to pay more attention to the impact of the exit strategy on the measurement indicators of a system itself.Therefore, you are better off judging the effectiveness of an exit strategy based on the most important metrics, rather than simply observing the changes after exit.

Also, when you think about your entry strategy, what you really care about is what happens after you enter the market, because entering the market is only the beginning of the game.Only by entering the market can traders make money.Exit strategies are different.What happens after you quit has no bearing on your results, only what happens before you quit.For the reasons above, you should judge exit strategies by their impact on the overall performance of the system.

(End of this chapter)