Swedish

Screenshots

Reports

A report in Optimal Trader will provide you with information about

• the latest trading signals
• expected returns
• win probability
• current settings

Click Report to show a report for your entire portfolio or your current equity. Do not forget to optimize the models before evaluating the latest trading signals!

With the help of the report you will get an estimation of the strength of the current models. You can base your decisions on two measures provided in the report:

• Expected returns
• Win probability

 

Returns

The returns (absolute or relative to a buy-and-hold strategy) are the average returns obtained when following the trading signals of the models. The values are calculated from a huge database of results from models with the same settings and the same optimization periods as your equities.

To get a better estimation, average returns are only based on model results when price patterns are similar to your equities current price patterns. The current price behaviour is classified in terms of

• the current trend strength,
• the current short term volatility,
• the long term volatility and
• Hurst Exponent estimation.

You can choose to analyze average returns relative to a buy-and-hold strategy or analyze absolute returns. This setting is available under Options / Best Models and Settings / Reports. It is important to understand that the estimated average return values only are valid under the following conditions:

• with the current optimization period and the current settings
• after optimizing the models
• with the current trading horizon
• with the current price behaviour

Please note that the results are an average based on results from a long period of time and from many different equities with similar price patterns. Variance is large.

 

Win Probability

Win probability is a measure of the certainty of the current trading signal. Win probability is based on an analysis of the price behaviour and the actual forecasting certainty from a large database of results from models with the same settings and same optimization period as your equities. To get a better estimate, the average (just like the average return) is only based on model results when price patterns are similar to your equities price patterns.

The higher the win probability is, the more often the model makes the correct forecasting. 100 means all forecastings are right, 50 means that the forecastings are correct half of the time, and 0 means that the no forecastings are right. A "right" forecasting occurs when the price rises after a buy signal or when the price falls after a sell signal. A wrong forecasting occurs if the price falls after a buy signal or if the price rises after a sell signal.

Win probability, that is predictability, is a measure that is easily misunderstood. A higher predictability does not need to indicate low expected returns. Imagine a stock that is trending with small price movements, but sometimes recoils strongly. The first 9 days the price rises with 0.3%/day. This will give us a profit of 2.7% under nine days. The tenth day the price drops 3%. A trend following indicator may have predicted the right direction in nine days out of ten, but still provided negative returns.

What is win probability saying us?

Win probability gives you an estimation of the certainty of the models trading signals. If a model has a low win probability, but high returns it may show us that the model is using situations not occurring so often, but which in the length can generate profit.

 

Trading Horizon

You can select between three trading horizons in Optimal Trader (one day, one week or one month) under Options/Reports/Best Models and Settings. The results for average returns and win probability are strongly dependent of the trading horizon used by the investor. Let us explain trading horizon with three examples:

Trading Horizon: 1 day

The investor optimizes his equities shortly before the end of the trading day and acts on the trading signals given by the models. He repeats this behaviour every day at the end of the trading day. The values of average returns and win probability in the reports are valid with this trading strategy if the trading horizon is set to one day.

Trading Horizon: 1 week

The investor optimizes his equities and acts upon the trading signals given by the models. After that the investor is passive for a week before he re-optimizes the models with the latest prices and acts upon the new trading signals. This behaviour is repeated week after week. The values of average returns and win probability in the reports are valid with this trading strategy if the trading horizon is set to one week.

Trading Horizon: 1 month

The investor optimizes his equities and acts upon the trading signals given by the models. After that the investor is passive for the next month after which he re-optimizes the models with the latest prices and acts upon the new trading signals. The behaviour is repeated week after week. The values of average returns and win probability in the reports are valid with this trading strategy if the trading horizon is set to one month.

Why are results better with a shorter trading horizon compared to a longer trading horizon?
The strength of signals from technical indicators diminishes with time. If you get a buy signal a certain day, certainty is larger that the price has risen a week after that day, than a month after that day. Much more may have happened in a month, diminishing the strength of the buy signal. The strength of a signal thus diminishes with time. A buy signal generated a year ago doesn’t say anything about the situation today.

At the same time shorter trading horizons are only better up to a certain optimal time period. When using trading horizons shorter than that optimal time period, results will get worse again. For many indicators in Optimal Trader it is better to evaluate the results of a trading signal after a couple of days than after a couple of hours. Indicators in Optimal Trader are not adapted for intra-day trading, but have their optimal trading horizon roughly between one day and one month.

Many investors do not want to be day traders, but are happy re-balancing their portfolio once a week. Although the reports may indicate that a trading horizon of one day is better, it is in fact not certain that this is the case when considering brokerage fees. Returns shown in the reports do not reflect brokerage fees.