The transaction model is a well-defined, step-by-step, rule-based structure used to manage trading activities. In this article, we introduced the basic concepts of trading models, explained their benefits, and provided instructions on how to build your own trading models.
Benefits of building a trading model
There are many benefits of using a rule-based trading model:
- The model is based on a set of validated rules. This helps to eliminate human emotions from decision-making.
- Models can easily be back-tested based on historical data to check their value before diving with real money.
- Model-based backtesting allows verification of related costs, so traders can see the profit potential more realistically. In theory, a profit of $2 might seem attractive, but a brokerage fee of $2.50 changes the equation.
- The model can automatically send mobile alerts, pop-up messages and graphs. This can eliminate the need for manual monitoring and operation. Using a model, traders can easily track the 50-day moving average (DMA) of 10 stocks crossing the 15-day moving average. Without this automation, it would be difficult to manually track even a stock DMA.
How to build your own trading model
To build a trading model, you do not need advanced trading knowledge. However, you do need to understand how and why price changes (for example, due to world events), where profit opportunities exist, and how to actually take advantage of them. Novice and experienced traders can start by familiarizing themselves with some technical indicators. These provide meaningful insights into trading patterns. Understanding technical indicators will also help traders conceptualize trends and make customized strategies and changes to their models. In this article, we will focus on trading based on technical indicators.
Simple trading model strategy example
According to the principle of trend reversal, some traders assume that falling things will rebound (and vice versa). Using the assumption of trend reversal as a strategy, we will build a trading model. In the following steps, we will go through a series of steps to create a trading model and test whether it is profitable.
Flow chart of building a trading model
1. Conceptual trading model
In this step, traders study historical stock movements to identify predicted trends and create concepts. This concept may be the result of a large amount of data analysis, or it may be a premonition based on accidental observation.
For this article, we use trend reversal to construct a strategy. The original concept was that if a stock fell by X% compared to the previous day’s closing price, the trend is expected to reverse in the next few days.
From here, look at past data and ask questions to refine the concept: Is this concept true? Does this concept only apply to a few selected high-volatility stocks, or does it apply to all stocks? What is the expected duration of the trend reversal (a day, a week, or a month)? What should be set as the downside level to enter the trade? What is the target profit level?
An initial concept usually contains many unknowns. Traders need some decisive points or numbers to start. These may be based on certain assumptions. For example, this strategy may be suitable for moderately volatile stocks with a beta between 2 and 3. If the stock falls by 3% and waits for the trend reversal in the next 15 days and expects a 4% return, then buy. These figures are based on the assumptions and experience of traders. Likewise, a basic understanding of technical indicators is very important.
2. Identify opportunities
In this step, determine the appropriate trading opportunities or stocks. This involves validating concepts based on historical data. In the example concept, we buy 3% on dips. First select highly volatile stocks for evaluation. You can download historical data of commonly used stocks from exchange websites or financial portal websites such as Yahoo! finance. Use spreadsheet formulas to calculate the percentage change from the previous day’s closing price, filter out the results that meet the conditions, and observe the pattern in the following days. Below is a sample spreadsheet.
In this example, the closing price of the stock fell below 3% within two days (February 4 and February 7). A closer look at the following days will reveal whether a trend reversal is visible. The price on February 5 rose by 4.59%. As of February 8, this change was lower than the expected 1.96%.
Is the result decisive? no. One observation meets the expectations of the concept (a change of 4% or more), while one does not.
Next, we need to further examine our concept on more data points and more stocks. Test multiple stocks at daily prices for at least five years. Observe which stocks have a positive trend reversal within the specified duration. If the number of positive results is better than negative results, then continue the concept. If not, please adjust the concept and retest or abandon the concept altogether and return to step 1.
3. Develop transaction model
At this stage, we fine-tune the trading model and introduce necessary changes based on the evaluation results of the concept. We continue to verify on large data sets and observe more changes. If we consider a specific working day, will the outcome of the strategy improve? For example, a 3% drop in stock prices on Friday will result in a cumulative increase of 5% or more next week? If we choose high volatility stocks with a beta value higher than 4, will the results improve?
Regardless of whether the original concept shows positive results, we can verify these customizations. You can continue to explore multiple modes. At this stage, you can also use computer programming to identify profit trends by letting algorithms and computer programs analyze data. Overall, the goal is to improve the positive results of our strategy and thereby increase profitability.
Some traders fall into this stage, endlessly analyzing large data sets with slight changes in parameters. There is no perfect trading model. Remember to draw a line on the test and make a decision.
4. Conduct practical research
Our model looks great now. It shows the positive profit of most transactions (for example, 70% of the profit is $2, and 30% of the loss is $1). We concluded that for every 10 transactions, we can make a considerable profit of 11 USD (7 * 2 – 3 * 1 USD).
Practical research is needed at this stage, which can be based on the following points:
- Does each transaction cost of the broker leave enough room for profit?
- I may need to make up to 20 transactions of $500 each to realize a profit, but my available funds are only $8,000. Does my trading model consider capital constraints?
- How often can I trade? Does the model show that transactions that exceed my available capital are too frequent, or too few transactions result in low profits?
- Whether the theoretical results meet the necessary regulations. Does it require short selling or long-term option transactions that may be prohibited, or may not allow simultaneous buying and selling positions?
5. Go online or give up and switch to a new model
Make a decision based on the results of the above tests, analysis and adjustments. Go live by investing real money using the trading model, or abandon the model and start over from step 1.
Remember, once you start using real money, it is important to continue to track, analyze, and evaluate the results, especially at the beginning.
6. Prepare for failure and restart
Trading requires constant attention and improvement strategies. Even if your trading model has been making money for many years, market development may change at any time. Prepare for failure and loss. Be open to further customization and improvement. If you lose money and cannot find more customizations, be prepared to discard the model and switch to a new model.
7. Ensure risk management by establishing what-if scenarios
Depending on the strategy chosen, it may not be possible to include risk management in the chosen trading model, but if things do not look as expected, it is wise to have a backup plan. What if you buy a stock that is down 3%, but there is no trend reversal next month? Should you sell the stock with a limited loss or continue to hold the position? What should you do if corporate actions such as rights issue occur?
There are hundreds of established trading concepts, and they are growing every day with the customization of new traders. To successfully establish a trading model, traders must have discipline, knowledge, perseverance and fair risk assessment. One of the main challenges comes from the emotional attachment of traders to self-developed trading strategies. This blind belief in the model will lead to increasing losses. Model-based trading is about emotional detachment. If the model fails, dump the model and design a new model, even if its loss and time delay are limited. Trading is about profitability, and loss aversion is built into the rule-based trading model.