Trading What I See

By the end of yesterday’s session, my two moving averages (13 sma and 34 sma) were intertwining and moving sideways, showing the lack of any trend. That’s one of the reasons I keep these simple calculations on each of my charts. Compare the first 5 1/2 hour non-trend with the last hour surprise.

During a trend, a pair of moving averages (of almost any length) will pull apart and come together as the market breathes. Impulse — retracement, far apart — close together, buy — take profits. But as indecision enters the market the moving averages tangle, and remind the trader that strong trends are the exception, not the rule.

This morning I watched price cross back and forth across the pair of averages, each pulse or pullback smaller than the last. It reminded me of the morning trading hours before an important Fed meeting. It also reminded me that sometimes markets will quickly set a high and low point, and then oscillate in smaller and smaller gyrations until the session close.

With this type of movement, it’s easy to get bored and miss a breakout on the few occasions it happens. Today was one of those days. There was actually a good entry on the pullback to the short moving average just before 12:30. While I was surfing the Internet. I’ll just remind myself of something I heard trader Don Miller say a number of years ago. “I don’t take trades during the last hour on Friday. If I made a dumb mistake then, I’d regret it all weekend.” Sure Don — it was only five points.

Today was a good example of why watching a longer timeframe is a good idea. It’s not to make entries, but to observe the context for setups on your trading chart. Tuesday I showed the way a longer chart can influence your trade timing in a negative way. This example shows how they can save you money.

The first 3-minute bar broke through yesterday’s high, and although price turned down within the 127% - 161.8% danger zone, the move looked pretty strong. The pullback stopped right at the .618 retracement just as the Stochastic hit the oversold area, and price was still above my longer-term moving average. A good setup. So why did I pass the trade?

Because on the 15-minute chart, that break above and then below yesterday’s high produced a Trader Vic 2B SETUP (see example.) The 2B occurs at trend reversals — in this case from up to down. My good setup is now in the wrong direction, and as you can see, taking it would have been a mistake. Instead I started looking for an opportunity to join the new downtrend and found two before lunch.

Someone asked me today whether using longer timeframes meant trading in the direction of the longer term chart. Not always. The trend is going to change first on the shorter timeframe, so the goal is to understand how your chart fits into the larger picture. In today’s situation the 15-minute chart showed an uptrend, but the pattern told me that below the surface the direction had probably changed to down. Even if this trade had worked out well, for me passing was the right decision. Good trading is always a matter of assessing risk.

There were several tradable moves at the beginning and the end of the day, but finding a few trades and then giving back the profits during congestion doesn’t put money in your account. The question becomes “How do you define congestion?” I do it by watching for overlapping bars.

In the first hour of trading there were two pushes to the upside, and in both cases the majority of the price gain came from bars that had little overlap. When this is happening, we can use the overlapping bars of a pullback as setups for continuation moves.

But by 8:00 (Pacific) we could see that conditions had changed. Starting with the pivot bar at 7:48, each of the next eleven bars had some portion touching the price level of that earlier pivot. Then we had overlapping congestion. By the time we broke away from one set of overlaps, we were already inside another. You can see this by the yellow horizontal lines on the chart.

How many bars does it take to create congestion? I don’t have a specific rule for that, but I’d say it should be at least five or six. At that point I become a very cautious trader. Notice that by the time you get five bars of congestion, you may already be in a trade. As you can see in today’s example, a long trade would have been profitable. It depends on your trading style whether you want to exit during congestion or to trail a relatively close stop.

I saw a nice entry early this morning based on yesterday’s market action. Linda Raschke calls it Three Little Indians, but my setup also requires a DIVERGENCE on the last peak and there wasn’t one on my 3-minute chart. But looking over at the 15-minute timeframe, there was the divergence I wanted.

So I entered the trade. From the 3-minute chart. And was stopped out. Oh well, sometimes a good trade takes a second entry. And sometimes the second entry gets stopped out too.

My trading plan doesn’t call for third entries, so I watched as the market moved down to my target. The losing trade doesn’t bother me nearly as much as having made a basic mistake — mixing up time frames.

Knowing what is happening on a time frame above and below your trading chart is important, but each time period has it’s own patterns. The longer the interval, the farther away the stop will be. This trade worked if taken on a 15-minute chart, but it would required a 3 1/2 point stop, which is more than I will accept while daytrading. But that doesn’t mean I can move the setup to a shorter period just to get a closer stop. And I know that.

Alan Farley, in his excellent book The Master Swing Trader, calls this a Trend Relativity error. “They see their trades in one time frame but execute them in another. This trend relativity error often forces a new position just as the short-term swing turns sharply against the entry.” Sure did, Alan.

I like to trade divergences, but I’ve been burned often enough that I normally require a second reason to enter a trade. Although today’s upside gap had some markets making a nice move, the Russell 2000 couldn’t exceed it’s first 3-minute bar. After the initial move up, the Russell and the SOX (semiconductors) lead the market down for the first hour.

As we approached yesterday’s low there was a clear DIVERGENCE setup and two reasons to trade — support from yesterday’s price action (the low), and the divergence. Divergence trades sometimes require more patience than other types of reversals, either by having relatively deep pullbacks or, as in this case, the sideways pattern between 8:30 and 9:30 (Pacific.)

But if you were able to wait it out, the move ended at a nice Fibonacci external retracement (white arrows.) The move from the bottom to the mid-session turn was 1.618 times the drop from the open. I always consider the range from 1.27% to 1.618% as a potential turn zone and use a close trailing stop.

By 10:30 we had a nice pullback to my longer moving average combined with a Stochastic oversold condition - two reasons to trade. On this kind of entry I don’t expect a deep pullback, so I exited for a loss. And didn’t re-enter on the divergence that followed. A stop below the first pivot would have kept you in the trade, but that doesn’t match my trading style.

The market’s rise ended at another common Fibonacci measurement — this time an extension rather than a retracement. Notice the general pattern for the day, where the first rally moves at a steeper angle than the second. When this happens, the second rally is often a 0.618 measurement of the first. It looks like it missed by two ticks.

Early trading moved quickly to yesterday’s high and reversed. Then there was a drop to yesterday’s low, with no pullbacks in sight. I suppose I could have shorted the early peak as a test-of-top, but I didn’t. Then yesterday’s low provided enough support to run us back to re-test the high.

By the time we reached the intra-day high again it was starting to look like a large rectangle would form. But at around 8:50 Pacific time we broke out. Or did we? There was a beautiful DIVERGENCE between price and the Stochastic oscillator that lead to the big move of the day. The reason seems to be a Philadelphia Fed announcement that was much worse than expected. The entry signal came a few minutes before the announcement.

But we didn’t know the news would be bad, so why would you really take the trade? Well, there was the divergence. And I just happened to be scanning several other markets at the time. Would it have helped to know that the NQ e-mini futures were making a lower high just then? What if I said that the ES e-min futures were just finishing a triple top? You say you want more? All making lower highs at that same moment were the YM e-mini futures, the NASDAQ Composite Index, and the SOX (Semiconductor Sector). Did I hear you say false breakout?

At 11:00 there was a nice pullback to resistance, combined with an overbought Stochastic, for a continuation move down. A trendline bounce gave you a nice exit.

Remember the context of your trade. Tuesday I mentioned the context of previous day’s trading pivots. Today it is the context of what the other markets are doing. Both are sometimes crucial to taking the correct action at the correct time.

One of these days I’ll convince myself to take Fed days off. I usually just watch and seldom make a trade. Today was no exception.

The morning started with a four point gap, followed by an eight point rally. With most commentators certain that Fed news would be No news, maybe today would be tradable. Wrong. The next three hours and forty-five minutes went sideways in a 4 point range. Then came the “excitement.”

Fifteen minutes before the Fed announcement, some traders tried to anticipate the news. Prices ran up to touch the daily high and then retreated. The report was released, and prices dropped to touch the daily low and then rallied. End result — we ended the day right at the center of the four point mid-day range. One of these days ….