Assuming breakouts at key levels are valid (let's not debate if they are or not) ie: support, resis, pivot, MA, ETC. What are some thoughts on the fraction or multiple of the ATR that the underlying has to exceed the level to confirm the breakout ? I thought I saw a post on ACD method where at 20-25% of the ATR over the opening range was considered appropriate. This seems pretty tight, except for the shortest time frames. I am thinking more along the lines of 2-5 day swing trades. Thanks for the input.
Hello sss12, The ACD reference is but one of many possible volatility choices to establish A levels above and below the high and low of a selected opening range (be it 5 minutes or 10 minutes or whatever one decides to use). One of my earliest mentors who regularly corresponded with me was/is John Hill from Futures Truth. He still trades ever day in his 90’s. His opening range breakout method (measured from the open and not some opening range) is 50% of the 10 day ATR. That is, on daily bars, once you have a confirmed breakout (on whatever BO criteria you're using) buy above the open plus 50% of the 10 day ATR. It tested as well as anything I tried.
Thank you. I've been using various % of the 14 day ATR (actual number very similar to 10), and 50% does seem to work well.
Now you're talking. You just utilized two concepts: your ATR was specified/couched in specific, parametric terms: "14 day ATR". This is a number specific to the underlying that you're trading (which is left out), AND to the time period over which the ATR was calculated ("14 day..."). So, you're half-way there. The second thought is that you wrote, "...and 50% [of the '14 day ATR' value] does seem to work well." which means that you're *testing* things. "There is no such thing as (real) confirmation." What works for this underlying will only accidentally relate to the next underlying you look at. And what works well for this week/month may only accidentally relate to next month's markets... ("Please!") Test test Re-test, and test again. Excise "confirmation" from your brain. Take a Bollinger Band (σ-oriented set-up) to your underlying -- translate that result to points of movement, translate that to an ATR look-back that equals the σ-derived points, and then play with the ATR on a going-forward, have-fun, trade-your-pants-off basis. But on regular occasion, RETUNE the whole business, so that it is empirically sound. And rather than using icky, ICKY words like "confirmation" get used to justifying wordy things like, "based on a sigma-derived/Bollinger-styled study of the immediate past historical stock behavior, I......" jus' thinkin....
I get what you guys are saying, I know there is no "real confirmation"and what may work today won't work tomorrow. I just wanted to get some discussion going on the topic, which it did, thanks.
It's not a matter of "what works today might not work tomorrow". To be a good trader, you need to focus on PRICE! To require a "price signal be confirmed by other parameters before you act" (such as volume, indicators, divergences, etc.) does you no good at all. Two problems... (1) the supposed "confirming" factors don't occur at once... they come along little-by-little depending upon their inherent drag/delay... all the while the trade is getting further and further away, and (2) regardless of however many "apparent confirmations" you see, it's all false. There are no genuine confirmations for any trade.
Get the no confirmation, I just acknowledged that in previous post. My original question/ discussion was along the lines of what margin of PRICE movement , as measured by ATR, above a BO level is indicative of a sustainable move ? Obviously, there are many other factors as well, but my question on this tread related to ATR.
Personally, I find no value in the concept of ATR.... just as there is nothing which is "indicative of a sustainable move". Of course you always hope a move is sustained, but the reality is that you have to "figure it out on the fly".