Trading What I See

We ended the day with the Russell 2000 right in the center of a large daily channel, and it also formed a candlestick Doji, ending just about where it started this morning. I think the market is confused. But at least it is forming tradable patterns.

We started with a drop, and then a quadruple reversal. Small gap below yesterday’s close, push up to match yesterday’s high, a move below the opening price, and finally a push to point “A” above yesterdays high, and then the real move down.

I wouldn’t try to trade any of that except for the last move. Notice that it formed a nice Spring pattern just like it did on Tuesday. A new high that can’t hold is usually a tradable move. I would have been more confident if it had happened with a divergence.

At 7:00 we made a bottom at the 78% Fibonacci retracement level, and then moved up for the next five hours. The up move followed one of the normal market patterns. When I start quoting myself, just remember that these are the patterns that can make you money. Here’s what I said Tuesday:

First pullbacks are often around a Fibonacci 62%. Second pullbacks are more likely to be 38%.

There were two good pullbacks in the five hour rise, the first being 62% and the second only reaching 38%. Now if only there were a way of telling when it was going to happen.

As long as I’m quoting myself, yesterday I said:

Although the frequency of occurrence isn’t quite the same, there are Fibonacci time measurements that often hit the same numbers I watch in price.

I mention 127% several times a week, as well as talking about the 100% Measured Move. Taking the distance from point “X” to point “B”, you can see on the chart my default settings for time measurements. The 62% retracement came at the first marker, and the 38% retracement came at the second. I believe it was Gann that said something about reversals happening when time and price come together.

Where did the day’s final top come? If you measure the distance from “A” to “B” and multiply by 127%, you end up at point “C”. To give a second indication that something may happen there, the distance from “X” to “A” is matched by the distance from “B” to “C” — a Measured Move.

A word of warning here. The market has a strong tendency to try fooling the maximum number of traders. Sometimes it does this by repetitive patterns that suddenly disappear. The last several days have really given some nice signals. Don’t be surprised if tomorrow is different.

If you’ve read this blog for very long, you know that, in addition to trendlines and channels, I’m usually looking for mathematical relationships — preferably several occurring at the same time. For example, I like trades where there is a 62% pullback that also contains a Measured Move to confirm a turning point.

But the relationships don’t have to limit themselves to price. Although for me it’s a secondary consideration, I also look for relationships in time. Today was a good example, but first lets look at the price setups starting with the morning gap.

Way back in 1932 Richard Schabacher (Technical Analysis and Stock Market Profits) wrote about gaps, and called the type we saw this morning a continuation gap. In 1948 Edwards and Magee (Technical Analysis of Stock Trends) gave a second name to this type of gap - a measuring gap.

As you can see, we’ve had a few years to watch this phenomenon, and somehow it just keeps repeating. Measuring or continuation gaps occur about half way in a move. Let’s add a little more information, this time from Victor Sperandeo (Trader Vic - Methods of a Wall Street Master.) Besides keeping these books within reach during the trading day, I have this Trader Vic quote on my desk.

If there is a gap, and it is going to reverse, it will do so 10 to 15 minutes after the opening 95 percent of the time.

This is a five minute chart — after the third bar, Trader Vic says the odds are in your favor. Can you find an entry to see if that was a measuring gap?

After the gap measurement was met we had a pullback. There was no reason to short there. Meeting a target is all about profit-taking, not about reversing positions. That comes at the next rally, when we had a nice divergence that matched a precise move to the 127% retracement of the pullback. I’m always looking for two reasons to take a trade.

Once the short sale was in position, there was no pullback large enough to cause concern until the 10:30 (Pacific time) bottom. And what reasons were there to change direction again? Lots of them.

Some traders like to take positions on filling 50% of a gap. I watch for this, but would never trade it in isolation. But in this case it matched a 50% retracement of the move that started yesterday. That’s enough to interest me, but there is something more involved.

Although the frequency of occurrence isn’t quite the same, there are Fibonacci time measurements that often hit the same numbers I watch in price. I particularly like trades where there is a 62% price retracement that takes 62% of the time of the original move. But I won’t refuse what we had today — a 50% price retracement that reversed at a 50% gap closure that took exactly 62% of the time of the rally.

As I’m sure I’ve said before, I really have trouble believing there is this much symmetry in the market. But that doesn’t keep me from trying to trade it.

After yesterday’s TREND DAY down, some sideways movement today was not a surprise. And once again, drawing trendlines and channels gave some structure to the movement.

We started with a small gap down, bounced for two bars, and then dropped again. There was always the possibility of two trend days in a row, but remember my basic philosophy — I look for places on my trading chart where a reversal might occur, and then wait for some sort of trigger entry. And this morning’s bottom was such a spot.

A week ago Friday I posted charts from several timeframes, including one where I showed a channel inside a larger channel, with all the trendlines parallel. I make it a practice to look carefully at a daily chart before the market opens. On November 16th we turned down from the outer channel. This morning we bounced exactly off of the inner channel. Show that one to the non-believers.

Does that mean we are going back up? I have no idea. I just try to trade what I see.

After our bounce off the trendline we made a 78% pullback. A key point about first pullbacks — they occur before the crowd believes there has been a reversal. That’s why they usually reach or exceed 62% of the first rise.

There were several ways to be sure you didn’t enter too soon. The first green Dunnigan bar was a possible entry. So was the break through the short moving average. Either way, the move carried right to the blue parallel, where you expect either a pause or a reversal.

When there is a second pullback more traders are convinced that a new direction is in force, so they buy earlier. First pullbacks are often around a Fibonacci 62%. Second pullbacks are more likely to be 38%. And that’s where this one stopped.

The next signal, whether or not you took the last rally, was what I introduced as a Wyckoff Spring. We broke to a high and immediately reversed, accompanied by a divergence in the Stochastic. But then notice how the move down forms a converging trend channel. To me that is a strong warning.

There were several potential bad reversal entries as we dropped through the lunch hour, but converging lines tell me to wait for multiple signals before an entry — like perhaps a nice divergence at the same time we hit a Fibonacci number. Have you noticed that the Russell seems to like 78% during congestion periods?

The rest of the day is a Measured Move, with the first rally matched by the second. If you caught the pullback, your exit was probably the parallel trendline which came a little earlier. Tomorrow we’ll find out whether the daily internal channel really holds, or if this was just a bounce in a larger downtrend.

After a few days away from the market, I tend to be quite cautious trying to find trade entries. This is particularly true after a light volume trading day as we had on Friday.

What I am looking for is an entry that tells me quickly (and inexpensively) when I am wrong. This is often a break through a level where a reversal would say “Get out now!” Sometimes it will be a moving average — other times a support or resistance level.

Fortunately, Friday set up a solid low to use as a benchmark. We opened down, but the previous pivot served as temporary support and we bounced, producing a very short-term double bottom. But it fulfills my requirement for a line in the sand, a place where I can have price action help make my decision to exit. But as you can see, it wasn’t necessary.

A fast, hard drop followed, and if you chose to take profits as the momentum slowed, you would still capture much of the total move. And by creating parallel channels there are several more potential entries for smaller, but still profitable, trades.

Point #2 pulls back to a steeply descending blue moving average. This is where I normally start looking for pullback entries. The fact that it moved slightly beyond the moving average gave another entry/exit point, the crossing of the average.

Price crossing a moving average, as a system, is a losing proposition. But that’s not the way I use them. At point #2 there is no question that the trend is strongly down, so all I’m looking for is a trigger that can be used in both directions — to get me in, and out if I’m wrong. A moving average can do that job well.

After the next pivot bottom you can draw a trendline, and of course, its parallel, which is hit at point #3. Once again we crossed the fast moving average, and once again it acts as a trigger.

Although a similar situation seems to happen later, the conditions are not the same. We eventually get to a parallel trendline and cross the moving average, but the momentum has vanished. You can see this as the yellow trendlines curve towards horizontal.

One advantage of drawing many trendlines is that it makes the loss of momentum obvious. And it takes good momentum to make this particular trading technique work.

While many people trade multiple markets or multiple stocks, when I am working on a very short timeframe I feel more comfortable concentrating on one. But that doesn’t mean that I ignore all the rest. On intraday moves, my market (the Russell 2000) sometimes leads the others, and recently it has been lagging. But overall, your best trades will occur when the markets are moving together.

How can you keep an eye on other markets without becoming distracted from your own? My solution is to standardize my charts so the important things stand out. I use two monitors during the trading day, but that doesn’t mean I am limited to two screens full of data. With a mouse click I can bring up a general market layout to see what is happening elsewhere in the trading world.

First the standardization. On my markets page each of the six charts has a green line showing yesterday’s high, a red line for yesterday’s low, a thin blue line at yesterday’s close, and two moving averages. All six of the charts are linked, so by simply typing in a number (3, 15, etc.) they will all switch to a different timeframe. The bars are color-coded (see Dunnigan post) so I can instantly see the direction of the last bar. And in the bottom pane is a modified MACD to show the overall trend.

In addition to the major futures markets (NQ, ES, and YM), I have a chart for the semi-conductors (SOX) and the two parts of the S&P 500 (the large cap and mid cap sections.) Although this is a screenshot from the end of last week, look at the information available at a glance.

The DOW futures (lower right) are above yesterday’s high (green line), while the midcaps (upper center) are still trading below yesterday’s low. All the other charts are inside yesterday’s range. Quite often, just watching the OEX and the MID charts, you can get a feel for the strength of a rally or decline. If they are moving together, there is momentum. If they are fighting each other, there is more likely to be congestion.

Seeing how price and the two moving averages are acting gives you another clue to trending or non-trending motion. Notice how easy it is to see that the SOX are the most congested at the present moment.

On a shorter timeframe, look at the colors of the last few bars. The mid-caps are rallying — trying to catch up (many green bars), while the other markets are going sideways (magenta or alternating red and green bars.) And the MACD at the bottom is oscillating in smaller and smaller moves as option expiration day expires in a sideways motion.

I keep this chart hidden to avoid being distracted from my interest in the Russell, but when I’m trying to get a feel for the markets, or just before initiating a trade, I’ll take a quick look at these charts. Never for signals, but for the context to know what to anticipate from my own trades.

Watching the overall market, the number of advancing and declining stocks, the rise and fall of sectors, and what is happening to volume is all a part of understanding the context of your trades. Although most traders can see the importance of this information, setting up a structure that is easy to use can be intimidating. I’ll be out of town the day before and after Thanksgiving, so I’ll use this opportunity to show how I track the rest of the market. I’ll start by talking about TRIN.

I have the actual numbers for the NYSE advances and decline, as well as up and down volume figures, showing on one of my monitors all the time (image on left.) But one key element of my trading requires that all my basic information can be interpreted at a glance, and the raw numbers don’t do that job.

Richard Arms came up with one solution in 1967. It is the ARMS index, but more people probably know it by the name TRIN, or the Short Term Trading Index. It is an attempt to measure the amount of volume going into advancing and declining stocks by combining their ratios into a single index. The number of advancing stocks is divided by the declining stocks to get their ratio. The same is done with advancing and declining volume. Then the first ratio is divided by the second to get TRIN.

According to Gerald Appel’s Winning Market Systems

This is a quite sensitive indicator and frequently changes direction moments in advance of market turnarounds on an intra-day basis. It is usuful in trading operations but its extreme sensitivity does result in false signals and fosters emotional activity.

He goes on to say that, in its raw form, the best use is to time the exact moment for pre-planned entries or exits in the market - not as a signal in itself.

There are many ways to use TRIN, and the most simplistic don’t work very well. You’ll find it written that a number less than 1.0 is bullish and over 1.0 is bearish. As a very general observation, this is true, but it is actually the direction (or change of direction) of TRIN that is most important to traders. And that leads to a visual problem, that in turn can lead to mistakes in a fast market.

When the market is rising, the TRIN tends to be falling. When you chart the TRIN under the market, a falling TRIN is bullish, and a rising TRIN is bearish. In one of his columns Dick Arms said that if he had it to do over, he would have turned the Arms Index upside down to make it easier to read. Of course, after almost 40 years, no one is going to change the actual numbers.

But with a computer you can reverse the direction of the numbers yourself. Because the indicator is so sensitive, many traders slow down the action of the TRIN, usually with some type of moving average. I use a modified MACD on the raw TRIN, and in the process, I flip it upside down.

The MACD (created by Gerald Appel - that’s his quote above) uses three moving averages to create a momentum oscillator. But by putting the numbers in backwards, you can reverse the direction of the oscillator. This way I accomplish two things — I slow the TRIN down to a level that matches my trading, and when my indicator goes up, so does the market.

The normal numbers for an MACD are 12, 26, and 9 — all exponential moving averages. The second moving average is subtracted from the first to give the MACD. Then you create a moving average of that result (using the third number) to create a trigger line. And if you want to reverse the direction of the oscillator, just reverse the order of the first two numbers.

Experiment with the input numbers to come up with a TRIN oscillator that matches your trading style. At the present time I’m use the Fibonacci numbers 34, 21, and 8 on various timeframes, but the goal is not to match my indicator — it’s to find something useful for you.

A magic indicator? Not at all. This is not even necessarily what your market is doing — it’s just another way of tracking whether your potential trade is riding the general market trend or fighting it. Have a nice Thanksgiving tomorrow, and Friday I’ll show you how I track multiple sectors of the market in real time.

Gerald Appel’s Winning Market Systems: Eighty Three Ways to Beat the Market

Yesterday had a small range, and today was an inside day (if you’ll excuse that two tick excursion at the close.) It was the narrowest day in the last 20. Looks like holiday trading has arrived. I set alerts to tell me when we got a breakout and they never went off. So instead, lets talk about reversal trading inside of ranges.

Just after the open we moved up to test a pivot high from yesterday, and missed it by a tick. We came back up and re-tested the same level forty minutes later, and this time went a fraction higher. Then we dropped to the Triple Crown bottom (see yesterday’s post), but lacked two ticks of reaching it. These are obvious tests of previous price levels, and I have them marked as Tests of Top and Bottom.

But when enough traders are trying to catch these reversals, their anticipation will sometimes prevent a true matching top or bottom from being made. Then you must make a decision on how close is close enough. If you get in too soon, you will have an excessively wide stop. And if you wait too long, you’ll miss the trade entirely.

Since volatility changes with market conditions, I like as many decisions as possible to be based on what price has been doing in the recent past — recent as in the last few moves I can see on my screen. So my definition of a Test is when the present bar goes anywhere into the range of the bar being “tested.” In the case of this morning’s bottom, I’ve drawn in the range of the Triple Crown’s lowest bar, with the high as a thin yellow line.

Sometimes I’ll fudge a bit, but this general rule keeps me from indecision when I should be taking action. As soon as price reaches the range of the previous pivot (assuming I’ve decided I like the trade) I’m looking for my entry instead of wondering whether we are “close enough.”

And speaking of the Triple Crown setup, the trade is still in play. Remember I pointed it out on a 15 minute chart. Derrik’s rules for this trade would have placed the original stop under the pivot as he took the entry on the next 15 minute bar. And he doesn’t move the stop until reaching his first target, which in this case is 796.90.

With the normal upward bias over the Thanksgiving period, another two points doesn’t seem unreasonable. Then he would trail a stop on half the position under the most recent pivot, with the rest protected at the entry level. For this trade he is obviously looking for a good sized move.