Forward Testing Single Prints
To follow the live day by day forward testing of the Single Prints please go
The theory that we are testing here in this strategy test is to determine if the single print provides a high probability area for us to fade the market movement. As with all strategies, we need to define success and failure according to certain criteria. What is a successful reversal point for one trader may not be for another trader because of the size of stops that each of them use.
For example, a strategy that provides 4 points of profit using a 2 point stop and reaches it's goal 60% of the time may only reach that target 10% of the time if used with a 1.5 point stop.
This is why we test the theories with money/trade management techniques: To provide us which as close to a mechanical system as we can get. Once we have a set of rules with a back tested probability we can work on ways to improve on that strategy or if time does not permit then we just trade that strategy.
There is another type of test that we can do which does not involve money or trade management. We can measure the probability of different sized moves (and their direction) after an event has happened. In this example that we are currently working with we would record the time, price and direction of the trigger event (the single print being touched) and then we would collect data over the ensuing time periods that record the size and direction from the trigger price that the market moves in each time period. We then use this sort of information to calculate stop sizes and profit targets which we then test further with money management testing techniques.
In this test I have not done any statistical testing of how the market moves after a single print is touched and here I am forward testing the single print as a trigger and a set of money management rules that I thought would be sound to run with this strategy.
Here are the rules that were set at the beginning of May 2005 and modified as appropriate during that month. These rules were used during this month of testing and it is anticipated that these rules will be modified during subsequent months in an attempt to improve the strategy.
- The single print must be less than a month old to be used.
- The single print must be a confirmed single print.
- The entry price is one tick inside the single print and an entry will only be recorded if the price touches the single.
- A stop of one handle will be used.
- 4 contracts will be traded.
- Partial profits will be taken at 1, 2, 3, and 4 points.
- Commissions will be $5 per contract.
- Stop will be moved to break even after first partial profit has been taken.
- Stop will not be trailed after moved to break even.
- No re-entries if stopped out but re-entries always taken if all 4 contracts are taken off profitably and re-entry is then only allowed on that day. (This is logical because single print will not be there on subsequent days.)
- All single trades will be taken irrespective of other news items or announcements that are being made at that point.
- If multiple singles created in same day or time period then all of them will be traded using the same stops and targets and rules.
- No overnight carry. Any open contracts are closed out at the last traded price (not the settlement price) that is traded during RTH which in this case is usually at 16:14:59 or 16:15:00 EST. Although you would not be able to implement this in real trading it appears to be the fairest way to measure this strategy. In reality you will close the trade before this time and get a slightly more or less favorable price.
- It is possible that we are running more than one strategy at the same time because we may have triggered an opposite direction single print trade while still trading a single print trade that hasn't reached its target when the next one was triggered.
- It is also possible that we are in more than one single print strategy trade in the same direction. All trades will be treated independently as if the other strategy was not running and the results for each recorded.