Two measures of buying and selling activity tracked on the Trading Psychology Weblog each day are the Adjusted NYSE TICK (a daily summed measure of number of stocks on NYSE trading at offer minus those traded at bid, adjusted for a zero mean) and the Institutional Composite (the same measure, but with the Dow Jones Industrial Average stocks). During the past five trading sessions, we've had four net selling days in the TICK and all five showing net selling in the Composite. I decided to investigate what happens after we get similar periods of significant weakness in both measures (average daily Adjusted TICK < -300; average daily Composite < -200).
It turns out that we've had 20 such occasions since 2004 (N = 708). Three days later in the S&P 500 Index (SPY), we've seen an average gain of .44% (15 up, 5 down). That is a meaningful bullish edge compared to the average three-day gain of .10% (397 up, 311 down) for the sample overall.
The market, of course, is on edge this AM awaiting the big jobs numbers. Keep an eye on the interest rate and currency markets to see if the news pushes us to new levels of valuation in those markets. If so, the news really is economic news and the adjustments of macro traders could lead to a sustained downward revaluation of equities as well. If the news doesn't really change the outlook for rates or the dollar, I'm going to question whether we'll sustain a major revaluation of stocks. In that scenario, the odds tell us that, as a whole, selling into a week of weakness is not a good bet.
Subscribe to:
Post Comments (Atom)
Posts
4 comments:
Brett,
What software/database do you use to find these types of statistics. It seems like very beneficial information that would be great to have at your fingertips.
Thanks,
~WilyTrader
Hi Wily,
Thanks for the interest. I archive my data from my real time feed and do all analysis in Excel. I do find the patterns to be helpful, although there are always exceptions to patterns. When multiple, different patterns point to a similar edge, very often there's a good trade there.
I'm getting so many interested inquiries about pattern testing that I'm thinking of conducting a seminar on the topic and teaching people how to conduct their own studies without a lot of heavy duty math or programming.
Brett
Hi Brett,
Thanks for a great blog. I too would like to do more analysis like you do on various sentiment and market factors. The problem I have run into is getting the historical data. I subscribe to Esignal but they don't seem to have very much history, especially on an intra-day basis. I do log on a go forward basis, but it will take time before I have enough data to make do a meaningful analysis.
Do you know of any sources for this data?
Keep up the great blog!
Jonathan
Hi Jonathan,
The Trader Development links on my personal site (www.brettsteenbarger.com) mention my main data sources. It is indeed possible to obtain good quality historical intraday data. I do think daily data is fine for getting your feet wet, and many portals (such as Yahoo! Finance) offer free end of day historical data for stocks, indexes, and sector ETFs.
Thanks for your interest!
Brett
Post a Comment