[atto]

In the Hinges thread, I posted a live trade and elaborated on my exit methodology. To keep that topic on topic and to allow us to continue investigating exits and scale-outs, I thought it would be a good idea to start a new thread. I'll cover my personal exit methodology seen though Wyckoff ideas, as well as the logical rationale behind it. That said, I welcome all feedback and supporting/opposing opinions.

My current trading methodology involves position adds and scale-outs. My reasoning has roots in wagering ideologies such as the Kelly criterion. The idea is simple: Bet more when you have greater odds. Relating to trading, you want the most exposure when your edge is greatest; similarly, you want less exposure when your edge is least (and ideally, no exposure when you have no edge). This makes logical sense, but many traders (myself included for a while) failed to see this.

Trading is a game of making money, not proving yourself correct. I disagree with the "A good exit is another entry" camp, because I can't say that my edge is always the same. Yes, if you're able to nail moves completely, then keep your all-in / all-out approach. I personally can't, and don't currently know of any trader who can.

Let's examine when a trading edge changes. Let's say you enter with a long setup (and many are discussed in this forum), and price moves in your direction but fails to break through a possible resistance area. Couldn't we argue that the new sellers, by confirming resistance, have taken (at least some of) your bullish edge away? This would be a good area to take some position off, because buying pressure is (momentarily) outmatched by selling pressure. I find volume especially important in these areas, because it can help you gauge the interest of the bulls and bears. Price stalled; did: a) sellers sweep in, or b) buyers simply take a break? If you see a rise of volume on the rejection, start looking for the door.

What if price did not stall at possible resistance? Then I see no reason to lighten the position. The buyers have been winning, and sellers didn't step in as they did before. In fact, I have position adding setups based on moves like this. Remember, price moves in waves (not bars) and is fractal. The setup you took on a 1m chart could parlay into a 5m setup. Always pay attention to the market on a greater scale.

My most reliable and accurate way to exit are climaxes. Here's how I define a climax:

1. A rise in momentium (volatility), along with:
2. A rise in volume
3. Then, a contraction of momentum / volatility (or, a rejection of price)

High volume does not mean there's a climax. In fact, some of the biggest moves are on high volume. It's the rejection / stall you're looking for. High volume gives you a head's up that there's a lot of interest. Once you spend screen time watching climaxes, you can catch them pretty quickly. Frequently, you'll see a quick decay of volume. This generally means that buying pressure has lessened, but sellers have not taken over. Many times, this is the making of a pullback before a continuation. If, however, you see volume gaining on the pullback, you might be looking at a reversal (or a pullback on a larger scale). It's not volume you're interested in directly, but volume's effect on price.

You'll also see times when volume does not spike before the exhaustion, but price fails to break through a support/resistance level. Many times, price will try more than once, but new buyers/sellers are simply not interested. This is another good scale out opportunity, because the lack of buying pressure is important. Price could likely continue, but our edge not as much as it was when we had buying pressure on our side.

So far, I've talked about exits that are pretty close to the extremes. Unfortunately, not ever price action move ends so cleanly with a climax or S/R confirmation. This is where stop management comes into play. To begin with, I use very small stops initially (so importantly, am willing to re-enter if my entry was not clean). Additionally, there's no reason to take a full stop if price is not confirming your entry premise. This is important. I am not saying to wait for the trade to be proven wrong. Rather, get out if you're not proven right. The Phantom of the Pits has some wise words on this topic.

So, we're in a profitable trade, and need to manage stops. My first goal is to make the trade riskless (move the stop to break even). This has many psychological and $ implications. Yes, at times, you can get shaken out for break even, and zoom!.. price shoots off. You must be willing for this to happen, and often, re-enter quickly without chasing a trade. I make the trade riskless as soon as price confirms my entry premise. This often involves a x point move, or a breaking of previous S/R. From there, I manually trail stops as price keeps breaking past S/R levels, or establishes new ones. Example: a bull run, and then congestion. I will set stops under the congestion. The more contracts you trade, the more scale-outs you can have, making your trade longer and longer (if this is wanted).

As a rule of thumb, the longer term the trade is, the less tight you need to keep your stops. On trend days, to catch the entire move, you will need to allow for pullbacks. In action, I'm frequently scaling out on pullbacks, and then adding to my position as the trend resumes.

This is a work in progress, so please feel free to add to my thoughts. My other posts on exits: Live trade exit discussion, the benefits of scaling out. A couple of you have mentioned to me that you like examples (helps solidify the concept). Please understand that this is simply one example, and does not represent the concept in entirety.

[Head2k]

Thank you for this topic. As a beginner building his first strategy, I find exits more challenging than entries. The reason is that while you can limit your entry to one known market situation or setup, once you are in a trade you must deal with whatever the market presents. Which might be much less familiar or obvious. That is also a reason I agree with the idea that the exit does not have to be a good place for entry in the opposite direction. Rather I think of exits as of changes in confidence.
I am looking forward to see this thread developing. Hopefully I will have something to contribute with soon

[Hlm]

Nice explanation atto.

I find it helpful to know where an example was taken. What instrument, time frame, and date was that taken from? Thanks.

[MidKnight]

Great post Atto. As you mention above, price moves in waves not bars, and I totally agree with that. However, volume consistently varies based on time of day. You will find that volume throughout the day is typically a smile. Because of this, I do not like to compare volume so much with prior volume at wave peaks/troughs to assess participation. Instead I prefer to measure the volume based on the degree of normal for that time of day. From there I believe I can get a better assessment of participation for that time of day.

I wonder how you assess probability as the trade is under way?

With kind regards,
MK

[atto]

Originally Posted by Head2k »

... once you are in a trade you must deal with whatever the market presents.... That is also a reason I agree with the idea that the exit does not have to be a good place for entry in the opposite direction. Rather I think of exits as of changes in confidence.

Good points. However, if you treat your exits as "exit setups", it gets a whole lot easier. Successful traders have given up the need to catch every move to its fullest extent. Often, I have to settle for a re-break of a previous S/R level for an exit (which means I gave up profits).

In chat, Hlm made the point that for his trading methodology, an exit is an entry, just on a smaller timeframe. In a way, I agree. A 1m climax might be a wonderful opportunity to make a counter-trend scalp trade. I don't trade that way, but in a way, my decreasing my position into a less favorable environment, I'm creating a trade in the opposite direction. Good stuff

Originally Posted by Hlm »

What instrument, time frame, and date was that taken from? Thanks.

I'd argue that it doesn't matter, because as I said, this isn't a representation of my entire point (possibly not even a good example). It just happens to show what I'm talking about. So just assume I made it up .

Originally Posted by MidKnight »

... Instead I prefer to measure the volume based on the degree of normal for that time of day.

I wonder how you assess probability as the trade is under way?

Interesting idea, never really looked into this. Please share if you have additional insight into this. I'm not huge on comparing volume on troughs and peaks, but it can show interest, which can help determine what the underlying supply and demand dichotomy is looking like.

My discussion of edge and probabilities is more qualitative than quantitative. That is, I don't have exact numbers for a setup (and when event x happens, how it affects my edge). Rather, I know what can affect my edge, and adjust my position accordingly. On a big climax (so, for example, very evident on even longer timeframes), I will be apt to scale more of my position out than on a small climax on my 5 or 10 second chart.


More on Climaxes
Something that's very important here is why climaxes "work". What is it about volume peaking, and momentum contracting makes for a good exit? It's actually very simple, and makes sense (which is important; you want your trading methodology to make sense logically). Climaxes show a fundamental shift in buying pressure vs selling pressure. Either buyers are content, are taking a break, and/or are taking profits OR sellers simply outmatch buyers. Either way, the peak of a climax is a wonderful exit, leaving time for the buyers and sellers to figure out who's going to win this squabble. Volume will tell you which of the two cases it is (in the above example, on climax I have highlighted, volume decreased on the pullback, and re-entered on the resume... this is telling that another push is probably under way).

Remember, climaxes frequently occur on very fast timeframes. Even if you trade off a 1m chart, venture into a faster time chart to spot these. It'll help on exit accuracy.

Also note: In my example, exits aren't on "bars". Bars are just cute summaries of continuous price data (which is why I'm a fan of a fast time chart; there's no different in "speed", just in horizontal displacement). Try not to get caught up in thinking what individual bars "mean". This became especially apparent a few days ago in chat, when a setup I took on a 10s chart looked completely different on another person's chart, because our bars were formed starting on different seconds. My setup / trade didn't change, but if I was relying on "bars", it would be vastly different.

(Cool trivia: Climaxes are often right after a WRB. Two different views of the market; the exact same result.)

[Hlm]

Originally Posted by atto »

In chat, Hlm made the point that for his trading methodology, an exit is an entry, just on a smaller timeframe. In a way, I agree. A 1m climax might be a wonderful opportunity to make a counter-trend scalp trade. I don't trade that way, but in a way, my decreasing my position into a less favorable environment, I'm creating a trade in the opposite direction. Good stuff

Exactly, especially with the following comment...

Originally Posted by atto »

I'm frequently scaling out on pullbacks, and then adding to my position as the trend resumes.

I also like the following comment of yours...

Originally Posted by atto »

I disagree with the "A good exit is another entry" camp, because I can't say that my edge is always the same. Yes, if you're able to nail moves completely, then keep your all-in / all-out approach. I personally can't, and don't currently know of any trader who can.

I completely agree. Going back to the discussion right above, many times I will scale out even though I don't get an actual signal on the smaller time frame in the opposite direction. One can have a setup (fractal profiling of price) without having a trigger (proper entry placement with required risk/reward). It's all about probability. The risk/reward to % of win ratio (positive expectancy) might not be within my range for the short, but that doesn't mean it's not high enough to hurt the probabilities of my long.

Originally Posted by atto »

I'd argue that it doesn't matter, because as I said, this isn't a representation of my entire point (possibly not even a good example). It just happens to show what I'm talking about. So just assume I made it up .

I completely understand. The only reason I ask is because I like to see how someone's strategy aligns with my own. But without having more data I am unable to compare. Maybe later you can throw up a couple of charts showing actual trades taken with the areas you scaled in and out...if you have time of course.

[atto]

Originally Posted by Hlm »

I completely understand. The only reason I ask is because I like to see how someone's strategy aligns with my own. But without having more data I am unable to compare. Maybe later you can throw up a couple of charts showing actual trades taken with the areas you scaled in and out...if you have time of course.

Definitely. A good example is a trade I took that I live posted, in the Hinges thread. Link to the trades, and then further down is my exit rationale. As this discussion develops, I don't mind adding other live examples.

[jonbig04]

Such great info! Thanks, I will be following along in this thread!