TrendMedium is a high performance tool for helping you get the best results
from your trading. You don't have to believe us, and in fact we encourage
you to prove it to yourself. Please free to try the tests and see how
they work for you. We're sure that you will not disappointed.
ē Walk Forward Test
Walk Forward Test is the most effective way to simulate a trading system being used on real data in real time. The test simulates the real life trading situation. It answers some very important questions about the trading system:
Walk Forward Test lets you verify the forward-trading
ability of the system. Now you can tell whether the system will continue
to bring profits after optimization or not. If your system is badly optimized
then you should expect it to perform poorly during Walk
Forward Test. On the other hand, if the system makes profit over
a large number of Walk Forward Tests, itís most likely to be successful
in the future. You can run Walk Forward Test
for a given number of last days and estimate the REAL performance of your
trading system on that data. Notice that simulation considers all your
settings for the selected analysis pattern.
|ē What this graphs show (Click on a graphic to enlarge it)|
These graphs show the sort of information that you can expect TrendMedium
to give you. In these tests we have used 15 minutes and 5 minutes time
periodicities to calculate the bars for NASDAQ. As you can see, TrendMedium
correctly identifies the turning points in the market and makes the decision
to buy and sell to give you a steadily increasing return.
|(Click on a graphic to enlarge it)|
These "forward backtests" allow us to check the performance of the model with known data, and to see how TrendMedium works in real life. Just imagine if you had this information at your fingertips in real time when you are trading.
Note that these results are not mock ups, but are the result of real test runs such as you will be able to do with your own data.
How TrendMedium does Walk Forward Test?
Prior to the test, the trading rules and optimal parameters are fixed. The system is then fed data from the same market that it has never seen before. This can be real-time data or historical data that did not form part of the data set on which the system was built. The results from the walk-forward test are then observed to verify their conformity to those observed during the in-sample test. A robust trading system will achieve a profit performance similar to that realised in the historical test, both in their size and distribution. It will generate approximately the same number of signals and display a similar risk - reward profile.
Hereís what happens when Walk Forward Test is run:
As you can see, itís very close to what you would do with paper trading.
Another feature of Walk Forward Test is that not only it lets you find
the optimal trading system, but also gives you the possibility to change
its parameters and see the effects. This way you can compare different
values for SUM, MOV ranges, and Calculation Period and choose the most
profitable ones. Please note that you can run Walk Forward Test with the
ReInvest feature turned either on or off.
www.trendmedium.com :: Eugene Labunsky 2002-2009