How to Trade Range-Bound Futures Markets: Tactics for Consolidating Sessions

The cost of a trend plan on a range day

Breakout trades are the classic way to lose money in a range. Price pokes through the high, you buy, the poke fails and price returns to the middle, and you're stopped out. Do that three times and the day is gone. I can't give you a trustworthy percentage for how often sessions range versus trend, because it depends on the product, the period and how you define the terms. What I can say is that you'll face both kinds of days, and it helps to have a written plan for each before the session opens.

Defining the range

You can't trade a range you haven't defined. One simple approach uses the first part of the regular session, which opens at 9:30 a.m. ET.

1. Mark the high and low of the first 30 to 60 minutes of RTH. These are your provisional boundaries.
2. Mark the midpoint between them.
3. Note any reference prices near the edges, such as the prior day's high and low, the overnight high and low, and any KPL levels you track.
4. Decide your rule for when the range is no longer valid, for instance a close beyond an edge by a set number of points.

KPL are fixed reference prices anchored to the RTH regular-session opening print, so the levels aren't complete before the open, and they don't move intraday. I treat an overlap between a range edge and one of those prices as something to look at. I haven't seen evidence that overlaps hold more often, so I don't size up for them. The KPL trading overview covers how the levels are built.

The edge fade, with numbers

The basic setup is to sell near the range high and buy near the range low. Here's a hypothetical ES example. The range is 5,400 to 5,420, so the midpoint is 5,410. Price tags 5,419.75 and stalls, and you short after a confirmation candle closes at 5,418. Your stop goes 3 points above the range high, at 5,423. That's 5 points of risk, or $250 on one ES contract ($25 on one MES contract), before fees and slippage.

The midpoint target is 8 points away, and the opposite edge is 18 points away. At the midpoint, the reward is 1.6 times the risk. At the far edge it's 3.6 times. Those ratios say nothing about how often either target is reached. To find the break-even win rate, divide risk by risk plus reward. At a 1.6 reward ratio, you need to win about 38 percent of the time before costs. Run your own numbers through the edge calculator.

Confirmation signals such as a rejection candle, falling volume on the approach, or a TICK divergence are common choices. I haven't tested any of them for you, and a signal that feels convincing in hindsight can add nothing. Pick one, define it so another person could apply it, and test it.

Scaling out at the middle

Taking part of the position off at the midpoint is a reasonable way to reduce the chance that a winner turns into a loser. It doesn't guarantee a profit, because the stop on the remaining portion can still lose more than the first portion made. Work out the combined result in advance. In the example, half a position out at +8 and the rest stopped at -5 nets a small gain, while the rest stopped at -5 after only a small scale-out might not.

When the range breaks

• Use a real stop order beyond the edge, entered with the trade, instead of a mental stop. Stops can slip in a fast break, so size for slippage.
• Don't add to a fade after price has moved through the edge. Averaging into a break turns a small loss into a large one.
• Be aware that ranges can resolve at any time of day. Afternoon volume and scheduled data releases are common triggers, so check the economic calendar and treat holiday and early-close sessions as exceptions to your normal plan.

Where the Marty bot fits

YMI's Marty bot is a separate automated product, and I won't describe its logic here, since current behavior and variants are documented on the Marty bot specs page. Bot access is part of the Pro tier, not Intro or VIP. Nothing about a range-day fade for a manual trader depends on it, and nothing in this article is a claim about how any bot performs. Models and strategies don't assure outcomes, and results can be worse in trending conditions or after a regime change.