Trading What I See

You’ve probably noticed that I try to point out turns that have multiple indications, since these setups are more likely to work out as expected. If three signals or indicators point to the same result, there could be three times as many other traders pushing price in your direction. Not everyone is looking for the same setups.

Today is an example of multiple versus individual signals. At the bottom marked “A” there were three reasons to expect a reversal at almost exactly the same point. At “B” there were also three potential reversal signals. But at today’s close I can see at least four different levels where a reversal would fit my trading style. And they are not the same.

At “A” we were making a double bottom with the pivot from June 8th. At the same time we finished an external retracement of 162% times the rise to “X.” That is often a Fibonacci turn measurement. And if you’ve learned how to use continuation gaps as a measuring tool, the distance from “X” to the center of the gap was exactly the same as the distance from the center of the gap to “A”. (For a refresher, see Trading Gaps from early April.)

Not only did the center of the gap mark the half-way point — it also provided resistance to the three-bar pullback after the gap. Each of those rising bars had less volume, so just from a chart-reading standpoint the first red bar could be an entry for a short sale.

At point “B” there was a 62% retracement of the entire move down from “X” just as we hit a trendline draw through the first pullback after the top. On June 4th I pointed out how forgotten trendlines drawn across first pullbacks will often mark later reversals — but you will only see them if you leave old trendlines on your charts.

Not shown (except for the Fibonacci level marked at the top) is an internal First Pullback Fib measurement. The top at “B” is at the “normal” 262% external retracement. That makes three reasons to suspect a reversal.

Contrast these pivots with the situation at the close. We have just hit the 127% external retracement of the rise from “A” to “B.” But if we fall a little farther, we’ll get to the yellow parallel. And if that doesn’t hold, there is always the long blue trendline that could turn prices upward. But before we get there it’s always possible that we’ll turn when “C”=100% of “A.”

Waiting for too much confirmation misses trades. But would you rather take a position at “A” and “B”, or take a chance at what will eventually become “C.” As soon as you see a potential setup, start looking for other indications of a turn. On good moves, it’s surprising how often there will be more than one reason to trade.

Today’s range just about overlapped Friday’s range — meaning there was little room to trade and lots of time to look at longer term charts. And last night I received an e-mail asking if these techniques really work on longer time frames. Let’s take a look at the chart I call my “Roadmap.”

For trading I use a three minute chart while keeping a 15 minute handy on a second monitor. But when I want to study the longer trend I use a 45 minute chart. As I’ve explained before, 45 minutes divides evenly into the 6 3/4 hours of trading for the Russell 2000 futures, giving nine equal bars for each day.

A Fibonacci measurement I pointed out yesterday was that very strong moves tend to reverse at a 423% external retracement of the first pullback. Notice the six day rally leading to point “A.” It’s a perfect hit of the 423% external retracement of the first pullback. Yesterday you saw it on a three minute chart — today on a chart fifteen times longer.

All of the other calculations work also, but first I want to point out something on this chart I don’t believe I’ve discussed in a commentary. A change of trend is often followed by a minor pullback before the move gains any strength. If you draw a trendline from a bottom reversal to this first minor pivot (bottom yellow trendline), it often seems to go off into space and have no relationship to price. By the time we reach point “B” most traders have forgotten that trendline.

You’ll find that point “B” is not a direct hit of a Fibonacci retracement from the earlier bottoms, but that trendline worked perfectly. That’s why I use multiple techniques in my trading. On the longer term charts I’m looking for potential reversal points — then I watch for setups on my trading chart.

Of course after a pivot such as point “B” I always tell you to draw a parallel trendline (in this case yellow.) Two days ago I showed how an upper trendline on a 3 minute chart pointed out the top of a move. On this chart that trendline is show in blue. And where the blue and yellow trendlines meet we have the completion of a large A-B-C with “C” equaling “A” for a Measured Move. Lots of reasons to anticipate the top that occurred on Friday morning.

Since that top on Friday we’ve moved sideways and are now bumping against the blue parallel. I don’t know which way we’ll go from here, but a combination of trendlines, Fibonacci, and a few chart patterns will help me find setups that continue to work. On any timeframe.

If you compare today’s chart with yesterday you’ll see almost the same pattern. There was a gap up followed by a rally and an A-B-C pullback that bounced from a natural support level (previous high instead of close.)

The pullback was 50% instead of 78%, but the same A-B-C Measured Move was there, and inside the larger moves there were smaller down-up-down patterns. And once again the parallel channel defined the turning point. At a quick glance the two day’s trading looks about the same.

The real difference was how to anticipate the intraday top. Yesterday it was by using a trendline extension on the 15 minute chart. Another way is by marking the external Fibonacci retracements of the first pullback in a move.

I check for external Fibonacci retracements every time a pivot is exceeded, but this is a special case. Often in an extended move, the first well-defined pullback will be retraced by 262%. Panic moves will shoot through the 262% level, but will usually run out of steam at 423%.

Sometimes these will lead to complete reversals; other times just to an A-B-C correction. As is often the case, this is not a trading setup by itself — just a great place to look for one.

So ends day four of a low volume pullback. Although I often remove the volume bars from the charts I show on the site for clarity, I watch it carefully during the day and especially as I analyze the longer time frames.

The longer we go without some good volume on an up day, the more likely it is that we haven’t seen a real bottom yet. Today the market just confirmed its indecision with another sideways narrow-range day.

Using some pivots from yesterday, there is now a large rectangle pattern that some may be trading as we bounce from support and resistance. In the Russell I want a rectangle to have a span of at least five points before considering trading a range, and I also want a divergence as it reverses at the top or bottom. Today we got one of each.

We started the day by trying unsuccessfully to penetrate yesterday’s high. Without a divergence or a close moving average I didn’t consider a short sale there, or a purchase as we moved higher at 7:00 (Pacific.) But when we came back to test the earlier top there was a good range and a nice divergence.

The same setup occurred in the opposite direction about 11:30. I wouldn’t take that first bottom, because there was no divergence. But the market came back down for a second chance at entry, and this time the Stochastic was moving up nicely.

The longer a rectangle extends the more likely a breakout, so if Monday continues sideways I’ll require more confirmation before taking another ping-pong setup.

For non U.S. traders, remember the United States begins daylight savings time this weekend, several weeks ahead of the “normal” schedule.

Sometimes Fibonacci measurements are obvious, and other times you have to make some tough decisions during the trading day. This morning was an example of how I attempt to make my choices.

The move up before lunch is marked with both magenta and blue A-B-C patterns. The magenta pattern started with yesterday’s 11:00 pivot low, and the blue pattern started at the magenta letter “B.” Looking back it’s easy to see that the blue A-B-C was the one to trade, but could you tell that in real time?

Remember that Fibonacci lines are just places where turns may occur — not where something has to happen. However if you have several Fib levels that match, a reaction becomes more likely. I had both patterns marked on my chart.

As soon as we exceeded yesterday’s high (green line), I drew the 127% external retracement of yesterday’s pullback (shown in white.) There is an almost perfect match between the blue A-B-C and that retracement, creating a Fibonacci cluster. As we reached that level, the Stochastic made a divergence with price, indicating a probable top and a good exit point.

The move down over the lunch hour also made an A-B-C pattern (not marked), but this one didn’t quite reach its target. The reason — it stopped precisely at the 38% retracement of the entire move. As is often the case, when you get close to a support or resistance level (or a Fib level), you must switch to price action for your final decisions.

The afternoon was another A-B-C pattern that once again ended at a 100% Measured Move. Notice that it was a flatter pattern than we had earlier. That often means that it’s time for either a pause or a correction. With tomorrow being Fed day, that’s not surprising.

Someone asked me last week for a good book on Fibonacci relationships. I probably have 10 or 15 that cover the subject in some manner, but the one that has influenced me most in the way I treat the A-B-C pattern was Forecasting Financial Markets by Tony Plummer. I just noticed that he has now expanded his book with new material in an updated version. If you are interested, see below.

Tony Plummer’s Forecasting Financial Markets

“Patience has Big Rewards.” That’s another note I have posted on the bulletin board right next to my trading setup. Take another look at yesterday’s non-trades. Fortunately I had the patience to ignore questionable signals. Today you’ll find the reverse — all my signals were there. And I had the reward of only taking signals that followed my style of trading.

The early rally ran up to yesterday’s high and reversed. That allowed drawing the first trendline from the previous high pivot up to point “1.” Whenever I draw a trendline, I try to draw a matching parallel, because that is often where the next trade will occur. Point “2″ is where price hit that parallel line, and it occurred precisely at a 62% retracement of the original move for the first clear entry of the day.

The rally from “2″ to “3″ wasn’t nearly as strong as I would have liked. For a good continuation I want the second rally to be stronger and faster than the first, so I was not surprised that the move ended between the 127% and 162% external retracement of the move from “1″ to “2.”

Now I have a new top pivot, which allows me to draw another upper trendline. And of course I draw the parallel. As I said yesterday, when the first pullback is at least 50%, I expect the second to end near 38%. When the 38% level occurs just as price hits my parallel line, it calls for another entry, but in this case a very cautious one. Here’s a case where my rules are met, but I’m not thrilled by the action so far. I really wanted that second rally to be stronger.

But with patience (and a very close stop) everything worked out as it should. The final top is a Measured Move, the distance from “4″ to “5″ matching the distance from “2″ to “3″. The divergence with the Stochastic was an added reminder that the move was probably in its last stages.

Is that the top? I have no idea. The “safe” parts of this short move are over for now. The last pattern looks somewhat like an upside triangle, but one with faulty volume. That means I’ll do as I always do — wait for the opening bell and re-evaluate as I draw more trendlines and Fibonacci measurements. And wait patiently for setups I really like.

Yesterday I mentioned marking the daily highs and lows on your intraday charts, and of course I should have included the previous day’s closes. That’s the dotted blue lines on my charts. You can see how they stopped price at the end of yesterday’s trading, and three times today before we finally broke higher.

Lots of traders use what they call pivot points - calculated levels from the previous day’s high, low, and close. Some even include today’s opening. If enough traders are watching one of the pivots they will work, but I prefer levels that everyone can see. Besides, by the time I include Fibonacci levels, trendlines, and reversal pivots I’ve got enough lines to keep me busy and find all the trades I can handle.

We almost made yesterday’s low after an opening drop, and there was a nice divergence there for a long entry. When your moving averages are moving sideways (notice the white 34ma), don’t be surprised if prices don’t quite make obvious targets. Traders get impatient and try to anticipate reversals.

Another thing that seems to happen when there isn’t a strong trend — the Fibonacci retracement of 89% seems to show up more often. And when an obvious target (yesterday’s close) matches the Fib level, better think about taking some type of action — in this case profits. The second hit of the closing level gave another nice divergence. A short here didn’t work out too well, because we were going sideways, but anyone taking it shouldn’t have lost any money.

A 62% retracement matching a Stochastic oversold condition is another nice signal, but with both moving averages going sideways I pass this type of trade.

What I’d like to point out is the last stalling action at the previous close (last blue arrow.) Alan Farley, in his excellent book The Master Swing Trader talks about this situation. When price comes up to a resistance area his suggestion is to short the second try and buy the third. He would actually buy in that hesitation before the breakout. Would have worked beautifully today.

Where do you exit. Once again, when a Fibonacci level (a 100% Measured Move) matches an obvious target (in this case a parallel to the lower magenta trendline), consider taking action. Prices could go higher, but you should at least take partial profits there.

Alan Farley’s The Master Swing Trader