Continuation Strategy, Fail-Safes, and Scaling

I’ve been looking for a strategy that trades more frequently. Looking at the m15 chart, I noticed that many mini-trends usually have a significant price move compared to previous bars, and additionally have many bars in the same direction at the start of the trend… i.e. it has strong momentum at the start of the trend.

So putting these two things into a strategy lead me to come up with the Continuation strategy, which is a kind of momentum strategy.

  • If at least ‘n’ bars are all in the same direction (trend)
  • And the total pip move of this since the first bar ‘s open in this trend exceeds a lower threshold, but is less than some upper threshold
  • Then assume the trend will continue, so enter a position.
  • Take profit can be defined as a % of the trend size
  • Stop loss can have intelligence enabled (Smart Stop) to be set at the low of the first bar (or high on a downward trend).

Additionally, this was further improved by closing positions and/or reversing positions when an opposite signal occurs.

After optimizing this, I found the following parameters yielded very good profit historically

  • Change Units:                         3 (pips)
  • Change Bars:                          48        (same direction is always true)
  • Lower Threshold:                    55 pips
  • Upper Threshold:                    68 pips
  • Trend Bars:                             5
  • Take Profit:                             185%
  • Close on Reversal:                 True
  • Allow Reversal:                       True
  • Smart Stop:                             True
  • Smart Stop Limit:                    1.0
  • Stop Loss:                               80 pips

The statistics for the above strategy were over the last 3 and a half years, trading a 40,000:

  • Win/Loss Ratio:                                   1.21
  • Expected Profit:                                  $524.56
  • Expected Loss:                                    $457.94
  • Expectancy Ratio:                              1.15
  • Total Profit:                                         $33,047
  • Total Loss:                                           $23,818
  • Profit Factor:                                       1.39
  • Min Equity:                                          $2,338


After doing some more research and speaking to a contact I know that works for a successful hedge fund, I learned about the idea of coding fail-safes into my strategies just in case there is a malfunction. Because of this, I have added the following safety checks

  • A daily limit on the number of trades
  • A daily limit on the max amount we can lose
  • A daily limit on the number of loss trades.

If any of the above are exceeded, trading is disabled for the remainder of the day. Additionally, all trades are subject to a MAX QUANTITY check to ensure that the notional can never be too high.

Multiple entries on same trade (scaling)

After discussions with a friend about the Panic and Retracement strategy, he gave me the idea that this strategy might be good to enter the position again if the original entry was pre-mature/badly timed. When we do this, we increase the notional (say double) to offset the loss of the first trade.

This is interesting but alters the way the strategies work. If we were to do this, that means we can no longer set stop losses on the trades as currently because we effectively need to carry this loss and hope the next larger trade will offset it. Of course, if the market does retrace then both trades will be profitable…

Or maybe it would be better to cut off the losing trade and just concentrate on the bigger trade.

Doing more research on this shows that the concept of scaling in/out of positions exists. This seems to offer its benefits depending on the strategy, so it may be something I can implement later. It definitely doesn’t need to be there from day 1 to make profit, but if you have a successful strategy, it can help bring in more profits and possibly cut back losses. Definitely something to revisit.

Regular tuning after optimising

I also got an interesting idea from my contact. Perhaps it is better to optimize over a short window period, say the last 12 months, and tune the strategy accordingly. After each month, this process can be repeated so that we re-optimize our strategy and update the parameters if necessary for optimal values based off recent history. It would be even better if this could be done automatically…


Posted in Fail-Safes, Momentum, Optimization, Research, Scaling and tagged , , , , , , .

Leave a Reply

Your email address will not be published. Required fields are marked *