Styles and Strategies for Scaling-In Trading
Yesterday we looked at scaling into a position. Today, we’ll look at what size the position should be when scaling-in.
If you decide to scale-in with 3 units (meaning 3 units gets you to 100%), you can do this equally, you can do this front ended, or you can do this back ended.
Scaling-in equally means you will scale-in for the same amount of money in each unit. Yesterday’s example used $10,000 per position. If you scale-in equally you will buy (or short) $3333 per unit.
Front end scaling-in means you will overweight on the front end. This could mean that you will use $5000 for the first unit, $3000 for the second unit, and $2000 for the final unit. Why would you do this? Because you believe there is a good chance that your first unit is going to be correct so you want to overweight it. And if you’re wrong, you’ll have cash available to scale in at better levels.
Back end scaling-in means you will underweight the front positions and overweight the back. One way of doing this is buying 20% on the first unit, 30% on the second unit and 50% in the third unit. The upside is that this is getting you a better average price (a lower average price) when you buy and a higher average price when you short when all three units are filled. The downside is if you nail the market on the first unit you will have a smaller position versus the two methods above.
In our testing (and we’ve run literally thousands of test variations on this), it appears that the back end scaling in is the best choice of the three. It quite often increases the times a trade is correct and it fits in well with the philosophy of high probability trading. But the results are not that much of a difference as to shift you away from the other two possibilities (For the sake of disclosure, we use equal position sizing in our e-mini trading on the short side. On the long side we back end scale-in.) and have for over the past two years. We may someday shift all this to a back end scaling-in approach, but we started it as equal scaling-in on the short side and it’s been successful for us. Again, they each have their merits and you should decide which one is best for you.
The most important thing here is to be consistent. Don’t jump around and change things in mid-course. Decide what’s the best approach for you and use it. And, if you’re looking for the simulated test results in a large universe of ETFs which have shown the highest percent correct, just know that back end scaling-in tends to be the better of the three. If you like high probability trading, scaling-in is a key ingredient.
Live Presentation Today: I’ll be doing a live online presentation today at 4:30 pm for our upcoming Spring 2009 Swing Trading College which launches this Friday. If you would like to attend the presentation, call 213-955-5858 ext 1 and they’ll reserve a spot for you. For more information about the presentation, click here.
If you missed Larry’s introduction to scaling-in trading, then click here to read “Scaling-In Trading.”
Larry Connors is CEO and Founder of TradingMarkets.com and Connors Research.