Monday, September 21, 2015

Trading Notes: Week of September 21st

Friday, September 25th

*  Once again we saw an expansion of stocks registering fresh lows on Thursday, as early selling dominated.  We saw 300 stocks across all exchanges make fresh 52-week lows, the highest number since August 24th.  Buyers became more aggressive into the early weakness and we closed well off the day's lows.  We've since rallied sharply overnight, consistent with the breadth query and model forecasts reported on Wednesday and Thursday.  My models remain moderately bullish over a 3-5 day horizon.  This will have me buying oversold weakness that occurs at successively higher price lows as a general game plan.

*  I've mentioned in the past that I like to track markets in event time, rather than in chronological time units.  Below is a chart where each data point for the ES futures represents 250 price changes.  This means that we draw more "bars" when markets are busy and volatile and fewer when they are quiet.  This 50-bar rate of change measure has been a useful gauge of overbought and oversold conditions per the game plan mentioned above.

*  I will be posting thoughts and ideas coming out of the weekend Traders4ACause event tomorrow and Sunday.  My talk will deal with research-grounded best practices for traders.

Thursday, September 24th

I'll be bringing a one-page best practice to the Traders4ACause conference this weekend and exchanging with others who choose to write up a useful trading strategy.  Great way to leverage mutual learning.  My practice will be an intermediate-term overbought/oversold indicator that captures both momentum and value effects in SPX.

*  We saw an expansion of new lows yesterday and again selling of bounces worked well intraday and again we stayed above overnight lows.  I continue to show us short-term oversold but intermediate term a bit overbought; my models are modestly bullish over a next three day horizon.  My intraday leaning is to buy dips that hold above overnight lows, but so far I have been less than inspired by the buying strength we've seen coming out of market selling.  A reduction in selling pressure is very different from an influx of buying: something I'll be watching going forward via NYSE TICK.

Wednesday, September 23rd

*  Once again, selling bounces worked well for much of the session on Tuesday as the decline continued.  We did see reduced selling and increased buying late in the session, but still closed with fewer than 10% of SPX stocks above their three and five-day moving averages.  When we've been in a moderate volatility regime going back to 2006 (N = 44), we've had 33 occasions up, 11 down for an average four-day SPX gain of +.83%.  Although we had weakness following the PMI number out of China, the market has since recovered and I would not be surprised to see further bounce from the short-term oversold condition.

*  I continue to find it useful to track the frequency of occasions in which NYSE TICK exceeds +800 and falls below -800.  It's been a good gauge of whether buyers or sellers have been predominantly in control and shifts in the distribution of those occasions has been helpful in identifying shifts from buying to selling and vice versa.

Tuesday, September 22nd

*  Selling bounces that failed below the Friday day session highs ended up being a good strategy in yesterday's trade, though we were able to hold above overnight lows and bounced strongly at the end of the session.  Action remains consistent with the thesis outlined yesterday that we put in an intermediate-term high with the Fed announcement.  My models are modestly bearish, overnight action in Europe has taken out those prior lows, and my intermediate-term measures continue overbought (see below).  We're short-term oversold at present, so my game plan is to sell short-term overbought levels that fail below today's overnight highs.

*  The intermediate-term strength measure takes a 10-day moving average of the percentages of SPX stocks making 5, 20, and 100-day highs minus lows.  (Data from the excellent Index Indicators site).  I generally like to be selling when new highs have been elevated but are now waning and buying when new lows have been elevated and are now drying up.

*  The measures of put/call activity that I follow, for all indexes and for all individual stocks, are on the low side.  I'm not seeing particular signs of bearishness on those measures, which has generally yielded subnormal returns over the near term. 

Monday, September 21st

Understanding how we best process information takes us a long way toward finding our edges in markets.  Most traders do not clearly understand and draw upon their signature cognitive strengths, in my experience.

*  We hit a buying crescendo after the Fed announcement, with NYSE TICK hitting multiple extreme positive readings.  What was significant was the strong selling pressure attracted by those higher prices, taking us lower late Thursday afternoon and pretty much all day Friday.  Friday saw 334 stocks across all exchanges register fresh monthly lows, highest in over a week.  My intermediate-term indicators remain elevated, and I'm operating on the premise that we put in an intermediate high with that Fed buying.  The quality of the buying attracted to the short-term oversold condition will tell a lot about where we go from here; my game plan is to sell bounces that fail to take out the day session highs from Friday.  I have one model neutral to slightly weak and another that is modestly bullish for the 3-5 day horizon.

*  Below we can see a chart of buying vs. selling balance since June; the data are derived from NYSE TICK.  The five-day moving average of this balance has been a useful short-term overbought/oversold measure.  Since 2012, when it's been above zero, the next four days in SPY have averaged a gain of +.09%; when it's been below zero, the next four days have averaged a gain of +.36%.  When the five-day average of buying pressure has been high, the next four days in SPY have averaged a gain of +.33%.  When buying has been in its lowest half of its distribution, the next four days in SPY have averaged a gain of +.03%.  When the five-day average of selling pressure has been light (little selling pressure), the next four days in SPY have averaged a loss of -.07%.  When the selling pressure has been heavy (top half of distribution), the next four days in SPY have averaged a gain of +.44%.