07/10/2021

How to test your trading system automatically

How Super Trend helps in testing a trading system

Super Trend is an indicator, which helps traders to identify buy and sell signals, based on the general market tendency. The indicator was invented and developed by Olivier Seban, a French financial consultant and writer.

We have already written about RSI, MACD, Bollinger Bands and Aroon. In this article, we will find out what the Super Trend indicator is and whether it is possible to use it efficiently in the financial market. Read in the article:

what a trend is;
how to add the Super Trend indicator to the chart;
what signals the Super Trend indicator sends (chart examples);
what results are provided by experiments on assessing the Super Trend indicator trading profitability in various markets;
conclusions.

Start using ATAS for free with no time limits! Or activate the advanced tariff right now to access the full range of functionality.

To try ATAS free of charge

What a trend is

A trend (tendency) is a situation when an asset price moves up or down steadily during a certain period of time.

Trends are important since traders, as a rule, make money when an asset price moves up or down. They make profit from buying when the price grows and, similarly, they make money by opening short positions when the asset price falls.

Many indicators try to identify a trend, so, how does Super Trend deal with it?

How to add the indicator to the chart and setting parameters

The Super Trend indicator is added to the chart in the ATAS platform the same way as other indicators are:

open the Indicator Manager in the chart window (or press Ctrl+I);
find ‘Technical Indicators / Super Trend’ in the indicator list;
press Add and OK.

The indicator draws 2 lines in the chart:

the blue UpTrend line (it is also called the support or Lower Line) is below the price;
the brown DownTrend line (it is also called the resistance or Upper Line) is above the price.

The lines alternate with each other. It is apparent that if the blue line is in the chart and the brown one is absent, the market is in the state of an uptrend, and when the brown line is in the chart and the blue one is absent, the market is in the downtrend.

Setting up the Super Trend indicator is easy. The indicator has 2 parameters:

Period (by default, Period = 14) affects sensitivity. The smaller the period, the more often the indicator will send signals.
Multiplier is a ratio, which takes into account volatility. The smaller the value, the closer the lines will be located to price fluctuations, and, consequently, the more often the signals will be received.

The indicator calculation formula is simple.

The ATR indicator, which measures volatility, is at its core.

Super Trend indicator signals

A trend changes when:

if a candle closes below the blue line, it is a signal of the downtrend beginning;
if a candle closes above the red (Примечание переводчика: может brown?) line, it is a signal of the uptrend beginning.

Example

Let’s consider the Super Trend indicator signals in the Brent oil futures chart (5-minute period).

Number 1 marks the downtrend beginning, namely on the candle closing at 16:05. There was a blue line ‘thrust’ on the previous candle at 16:00, but it closed higher, that is why there was no signal.
The downtrend developed and a wide bullish candle, which thrusts through the blue line, appeared at 17:30. The closure, however, was lower, that is why the downtrend continued.
The uptrend beginning signal was received only on the candle closing at 17:35.
The blue line ‘withstood’, since the candle closed higher.

Testing the Super Trend indicator trading profitability

Is it possible to make money trading on the exchange by the Super Trend indicator signals?

To answer this question, we have developed an algorithm, which collects the Super Trend indicator signal statistics, and we have connected it through API.

The How to write algorithms for ATAS article contains an easy-to-understand description of how to write your own algorithms in the C# programming language for novices.

The algorithm has a simple code:

Calculation of the Super Trend indicator trading results

The algorithm uses default settings and goes through all candles in the chart.

IF:

there is a blue line on candle ‘i’
and there is no blue line on candle ‘i-1’,

THEN:

we close the short and note down the result in the ‘profitshorts’ variable
and open a long.

It is similar for sells. The picture below shows how the long closes and the short opens at the ‘i+1’ candle opening price.

When the algorithm reaches the chart end, it displays the final result in logs.

Experiment 1. We add the algorithm to the gold futures (GC) Continuous chart (the data are from the CME; period = 5 minutes).

Let’s see what happens in the log window.

Подсчет результатов торговли по индикатору Super Trend

The algorithm has calculated that the following was done by the Super Trend indicator signals during 111 days of trading 1 contract:

874 sell trades were executed, and the total result of sells was -83.6 (USD -8,360 when trading 1 contract);
875 buy trades were executed, and the total result of buys was -29.9(USD -2,990 when trading 1 contract);
the average result per trade was -6 ticks.

If we add a commission fee (let’s take USD 5.5 per cycle at an average), it turns out that the commission would have cost approximately (874 + 875) * 5.5 = USD 9,619.5. It is a bit disappointing.

Let’s try to change the time-frame / chart period / chart type. What will happen then?

Experiment 2.

Let’s add the algorithm to the E-mini S&P 500 futures chart. The data are from the CME; Range (8) chart; Scale = 4. The chart covers 230 days.

We can see the result in the log window. 598 trades were executed by the Super Trend indicator from October 7, 2020, until August 26, 2021 (the commission was approximately USD 3,289). The average trade result was a bit more than +1 tick without taking into account commission fees and slippages, which means that the approximate result was close to zero.

It is interesting that the total result by longs was USD +33,050 or 2,644 / 299 = a bit more than USD 110 per trade. As regards the sells, they brought losses. It is explained by the fact that the US stock market developed exponentially during the period, which was loaded to the chart, against the background of unprecedented support from the authorities.

Does it turn out that the Super Trend indicator can be used with a small effect for taking profit on ascending sections within the framework of global growth? Yes, it does, but it is also true for many other technical indicators.

Moreover… If we load 30-minute data for a period of 1,000 days (from the end of October 2017), we will get the following data from the algorithm:

If we execute 2,463 trades, we will not be able to make profit. Each trade will bring an average loss of 1 tick or USD 12.5.

But what if we take buys only? As a matter of fact, the S&P 500 was 386.36 at the moment of its opening on January 1, 1957. It reached 4,500 in August 2021, which means 11 times growth in 64 years or approximately 7.5% a year.

The algorithm shows that if you open longs by the Super Trend indicator during 1,000 days, you will be able to make 3,680 ticks * USD 12.5 = USD 46,000 (with consideration of fees: USD 46,000 – USD 6,776 = USD 39,224) having executed 1,232 trades.

However, the index was 2,575 at the end of October, and if we had opened a long and held the position, the profit would have been 4,500 – 2,575 = (1,925 * 4) * USD 12.5 = USD 96,250. It is more than 2 times compared to the buy signals from the Super Trend indicator. No comments.

Experiment 3. Let’s have a look at the Renko (6) chart in the gold market.

The algorithm has calculated that the following was done by the Super Trend indicator signals during 111 days of trading 1 contract:

2,452 sell trades were executed and the total result of sells was USD 54,570 (USD 41,084 with consideration of fees);
2,453 buy trades were executed and the total result of buys was USD 60,200 (USD 46,708.5 with consideration of fees);
the average result per trade was USD +230.

Both buys and sells bring profit. Why did the result improve? It is due to the fact that Renko charts process market data in a different way. The more pronounced a market trend, the more frequently candles appear in the Renko chart and more rarely new ‘bricks’ appear in a flat, which can reduce a number of false signals.

However, there are 2 nuances.

First nuance.

Look at August 8 (a ‘wide’ volatile day in the chart). It was the Flash Crash in the gold market. The price rushed down with acceleration, as a snowball, under conditions of low liquidity on Monday morning, when only the Asian markets had opened. It took place in the context of the negative US unemployment news, which was published on Friday. Trading robots and closing of longs by stop losses also added to that.

The algorithm calculated that day as if trades were immediately executed at those prices, which were topical at the moment of the ‘i+1’ candle opening. However, in actual practice, there would have been terrible slippages and if, in theory, this day shows a profitable performance, in reality the result could be negative.

Second nuance.

It has to do with how ‘brick’ openings are built when the ‘bricks’ change their color. If a ‘brick’ in the Renko chart changes color, its opening is built on the previous candle closing and, since our algorithm calculates effectiveness by the opening prices (immediately after a signal has been received), it brings some uncertainty into the result.

How strong is this uncertainty? Judge by yourself.

If we add the algorithm to the Renko chart (12) for the Bitcoin market on the Bybit exchange (Scale = 25, since the price step on cryptocurrency exchanges is very small), we will get a suspiciously optimistic result:

1,127 trades were executed during 112 days and the average result per trade was nearly plus USD 90 (when trading 1 Bitcoin), and the total profit exceeded USD 100 thousand.

However, if we switch to, for example, the Volume = 500 (Scale = 25) chart type, we will get the following picture:

1,097 trades were executed on the horizon of 114 days, and each of them brought a bit more than USD 29 without consideration of fees. The total profit decreased more than 3 times with a comparable signal frequency in the same market. It became approximately equal to the commission fee size.

Conclusions. If, while testing on history, you receive highly profitable results, perceive it with a degree of scepticism and conduct an additional test. Paraphrasing a popular proverb, ‘Measure thrice on demo and cut once on real’.

Using the Super Trend indicator with basic settings in the popular markets seems to be a futile idea. It can make a rational sense, if you use the indicator in a non-standard fashion or in a combination with other instruments.

Advantages and disadvantages of the Super Trend indicator

The Super Trend indicator has the same advantages and disadvantages, which are inherent in many other technical indicators that are built by simple formulas on the basis of the past prices.

Advantages:

it allows to gain benefit from long trends;
it is not redrawn;
it is easy to understand and use;
it provides a small statistical advantage when settings are adjusted.

Disadvantages:

a possible profit from using the indicator is comparable with the size of commission fees;
some signals come with a delay;
a big number of false signals come in the flat markets.

It is hardly possible to eliminate all disadvantages of the Super Trend indicator.

However, try to use the following in order to smooth them down somehow:

Combination of periods when signals from senior and junior time-frames coincide.
Technical analysis. Open shorts only near resistance lines and open longs only near support lines.

We also recommend that you pay attention to:

cluster charts, since they contain valuable information, which will help you to understand what really takes place in the market better;
indicators, which are based on the order flow, such as Imbalance, Cluster Search or Tape Patterns.

Conclusions

The Super Trend indicator is a simple instrument, which is calculated on the basis of the current market volatility, which is measured by the ATR indicator. The indicator plots 2 lines in the price chart. If a volatility splash results in the price growth and crosses the resistance line, the buy signal is activated. It is similar for sells.

The statistics collection experiments show that the Super Trend indicator can hardly be used as a self-sufficient instrument.

Nevertheless, we accept a scenario, under which you:

carried out a trustworthy study and found optimum settings for the indicator in a certain market;
tested it on your demo account and made sure that the indicator really helps you to make profit;
feel confidence and comfort in working with the Super Trend indicator…

… – in this case you have reasons to use the Super Trend indicator on your real account.

Have you ever tried to trade by the Super Trend indicator? Share the results you managed to achieve.

From our experience, we would recommend you to develop your cluster chart reading skills and rely on more progressive instruments, which are oriented towards volume analysis.

Download the ATAS platform free of charge and check how its indicators and cluster charts can really help you.

Strategy file

Information in this article cannot be perceived as a call for investing or buying/selling of any asset on the exchange. All situations, discussed in the article, are provided with the purpose of getting acquainted with the functionality and advantages of the ATAS platform.