Tuesday, 13 September 2011

Why Double Down

Doubling Down or averaging a losing position is strongly criticised by many people. I'm not one of them.

I'm not advocating to average every loser, but doubling down is a valid strategy to have in your arsenal of tools to use as a discretionary and as an algo trader.

A good example is from yesterday's ES chart.

The ES opened in RTH and I switched to the Dow Euro after I had a couple of nice longs in the ES. The ESTX50 trades a little differently from the ES as it's a European stock index.

I sold the Fib, as you can see on the chart, looking to buy back as it came back into the Keltner Channel towards the 33 EMA. It didn't happen as expected and I sold twice more as it powered through the Fib area and went overbought. By that time it was way too far from the 33EMA and I was confident of a pullback as eventually it always comes back to the 33EMA. I did have a drop dead stop.

I ended up with a 2020 average sale price and took 5 points out of the trade on the first push down. I covered half there as the price had broken through the Fib area and was looking for the price to then bounce and either take off or fail. We were far enough from the 33EMA for me to believe I could take a profit on the second half of the piece. In fact I was able to re-short the half that I had covered as the breakout failed, covering nearer to the 33EMA.


  1. This is a "rule" that I break, also. The emphasis for this rule is certainly for those that enter a full position, then continue to add to that position as it moves against them.

    But if you are entering a partial position in multiple entries over a range of price, until you reach a full position, then there is no difference between this and one entry, except for a few extra commissions.

    Assuming, of course, that you're practicing good risk/stop management for both methods. If you're not, or you're not comfortable with the DOM, or you don't have a statistical understanding of the indicators that you use, then "don't double down" is a rule to follow. Then again, "don't trade live" is probably a better rule at that stage. :)

  2. Could you please tell us in this example at what price you did put the drop dead stop for the averaged positions? what is the %of account for the drop dead stop?

    I use a similar strategy for 10% max draw down and max 3 averaged position, so It would be nice to know your numbers.


  3. Mr. EL,

    In your Aug 29th blog you demonstrate a 1.25 Renko bar algo using TradeStation. A word of warning: TradeStation Renko Bar backtesting is completely unreliable. You can set your stops at brick size plus one tick and always get a good result. Even the walk-forward results look great. Unfortunately, the algo will perform poorly in real time. The problem lies in the way TradeStation packages its Renko Bar data.

    Best regards,


  4. Anon 18:42, None of the renko backtest results are worth anything as is, due to the fact that the wicks are ignored. It requires a number of adjustments to inputs to make it useful in real time. But thanks for the heads up.

  5. Alex, my drrop dead stop is usually $400 per contract except for the DAX when its 500 Euros. I move my stops in line with market context ance the trade is underway. My drop dead stop is never hit unless I have an internet issue and that hasn't happened for a long, long time.