Wednesday, July 15, 2009

Using Stochastics On Multiple Time Frames

Click on Charts to Enlarge

This post is just for educational purposes. It will be showing stochastics on several time frames paying attention to a extreme readings on several time frames occurring at the same time. The top chart is a daily chart of QQQQ. Of note is that this morning the RSI(3) is above 90. I have highlighted past instances with light blue vertical lines. What we see is that this has typically occurred just prior to significant peaks in the market with the downside risk being large, and upside potential being very minimal.

Also, I have put a small pink line showing that the gap down from July 2 has been filled as of this morning. From a swing trading perspective, in this case looking to go short, I like to see overhead gap downs filled to remove some risk due to tendency for gaps to fill within the first few weeks after forming.

The charts below are of SPY, not QQQQ, but they all look similar. The next chart is a 60 min chart with stochastics underneath. We see that it became overbought yesterday (pay attention to the red line). Now the blue line is actually lower than it was the last couple days despite the large move up today. So this is a first hint of divergence on this time frame.

The next chart is a 15 min chart of SPY. The stochastics underneath is overbought on this time frame as well. Additionally, while price has made significant new highs the last 2 days, the stochastics peak (red line) is only at about the same level. So while it is not clear yet, this may end up showing somewhat of a divergence.

The bottom chart is a 5 min chart of SPY. Again the stochastics is very overbought. You could also look at a 1 min chart as well, though that would only be at all necessary for a day trade.

The point of this is to show that when cycles from multiple time frames are all extended it is often possible to enter based on a short-term chart like a 5 or 15 min with a narrow stop based on that time frame, but it may be possible to hold for a longer swing trade for the exit (exit with the 60 min chart or daily RSI(3)) if you get immediate confirmation and don't get stopped out. It is certainly possible to get risk/reward ratios upwards of 10 to 1 at times like these. Now granted you may get stopped out a few times, but you only have to be right 1 out of 10 times to be profitable in the long run.

In this particular case, I would probably wait for the 5 min chart to come down and then give an overbought signal at a lower high befor entering for a potential swing trade. Then set the stop above the high of the rally.

Now, it is not worth beating a dead horse on the current SDS blog trade. At this point, the best course of action is to simply wait for the exit signal to come and take the result win or lose. The market is extremely overbought short-term. I can't remember the last time the short-term model was this overbought (maybe January of this year, but I'd have to check the data). Also cumulative intraday TICK levels are thru the roof. So either the market will put on the rocket boosters and go to the moon, or more likey, mean reversion will kick in over the next week or so and bring prices back down a good bit. It has been a rare past blog trade where the market makes this much headway against the trade before reverting. In those cases, when the exit signal comes, it tends to result in a near breakeven trade, maybe with a small loss (small relative to the average winning trade). It is still certainly possible that this trade ends up with a profit, but it would probably take a decent/large size gap down or two before the exit signal, to be able to have anything more than a small win.

Now on a side note, the middle of this week should be a cyclical turning point. In a recent post I had suggested/projected that it may be a low point before a substantial rally. However, the move up since then makes it more likely that it will be a cyclical high (if this time frame does prove to be an inflection point). While it will only be relatively clear in hindsight, currently I believe the best two interpretations of the current price pattern both suggest very substantial downside for the next 1-2 weeks.


  1. I have mentioned this before on the blog Jan 2 2009 and Feb 6 2009 about when the market makes a large up day, especially if it closes well above the open, and the VIX makes a hammer reveral candlestick.

    While it is still mid day as I type, the cash VIX is up slightly for the day and is way off the lows. If this configuration continues into the close, I have noted it being a good bearish warning. Basically there is recognition in the options premiums that the market it too stretched and the premiums don't fall as the market rises. During this bear market, this has foreshadowed some large mutli-week down legs.

  2. Today's reading for the short-term model is the lowest that I can find history for. The last comparable readings were January 5th of this year which was the day before the high of that rally before rolling over severely for 2 months.

    The prior time was July 11 2005, which was about 3 weeks prior to a 2+ month decline. Even in this case the max gain was less than half of the max loss over the next couple months.

    Other comparable instances not quite as extreme were Dec 26 2007 and May 26 2006 which both led to sharp declines over the next couple weeks.

    I think today (or maybe tomorrow) will mark a significant peak and these levels may not be seen again for several/many months......or I could be on crack.

  3. i hope you are not on crack. i'm probably going to add to my i-etf positions today. just a few more bears to give up, and hopefully we'll move down. to me, the upside risk seems so "risky" here when compared to the downside potential... blah, blah, blah... . you do that analysis so much better than i ever could pete... (smile)


  4. People have asked me before about adding to losing positions. With a mean reverting system I think it can indeed be a sound strategy. However, I would only suggest doing it if you planned it ahead of time by taking a smaller than normal position at the initial entry or your general allocation to each trade is fairly low, such that by adding, you don't get out of hand.

    With that said, there are several signs today that this should basically be "it" for this move. The bigger question (in my mind at least)now is how far down the next leg goes. I would say it definitely comes back down near or just under the recent lows at a minimum before any tradable bounce back to the upside.

    I will probably have to play the next few weeks pretty mechanically for trades because price is thus far rangebound and I don't think it can breakout to the downside without becoming oversold short-term again.