I'll be blogging and tweeting today from a proprietary trading firm, where I'm working with traders. Stay tuned, and I'll pass along insights during the day. You can read the latest five tweets on the blog page under "Twitter Trader" or you can follow the tweets on my Twitter page.
So how do I work when I'm at a prop firm? I roll into the office at 6 AM, well before anyone has arrived. I spend the early morning running historical studies that will provide insight (and perhaps some edge!) for the traders I work with. As a coach, I want to be able to speak knowledgeably not only about traders' psychological states, but also the states of the markets they're trading.
Most of the historical studies I research and share with my traders don't get posted to the blog. It's my desire that the traders I work with derive unique benefit from the research efforts. If I've done my work well, I've earned my day's keep before the opening bell has rung. Just one good idea to a trader who trades 200+ S&P e-mini futures at a time can provide a meaningful return.
For today, however, I'll share one of the ideas that I'll run by interested traders at the firm. On Thursday, we traded below the S3 support level on a weak day. Today, we're poised to open below the S1 support level in further weakness. When the S&P 500 Index (SPY) trades below S3 and then opens the next day below S1, what has happened historically from the open to the close in the coming trading day?
It turns out that such bearishness does not bring a bearish edge to the trading day. Since 2000, we've had 92 instances in which the market has followed a very weak session with a weak open. During the day session, the market has averaged a gain of .25% (51 up, 41 down). By comparison, the average day session change for the remainder of the sample has been -.03% (1149 up, 1138 down).
Recall that my previously posted research anticipated a lower market following Thursday's weak momentum. Now we have to temper that finding with the above research. Should we have difficulty breaking and staying below the overnight lows in early trade, I will entertain the possibility that we could see a bit of squeeze on the shorts. At the very least, I'll be encouraging traders to not automatically assume that recent weakness will bring a bearish edge to today's day session.
.
Subscribe to:
Post Comments (Atom)
Posts
5 comments:
Thank you as always Dr. I always amazed by your hard work and work ethic.
Thank you for this interesting and unique insight regarding historical data. So as the resident psychologist among other traders, do you trade at several different firms, or just at one firm?
Keep up the great posts!
Dr. Brett, your work is good. And I'm sure you use other indicators too.
Not to be pushy, but you really may want to study the efficacy of the ChargsEdge "daily maps" that I have the privilege to also post at my site, they are also normally available freely each day before the open at www.Chartsedge.com/wp (where they are generated by Mike Korell).
They work well enough to take out lots of the intraday anxiety of looking for high/low turning points. Which I know from personal experience, is a huge benefit.
Paul and Peter play the proverbial "toss a coin" game betting 1$ on each round. After 100 throws Paul is ahead by 20$ (60 wins / 40 losses).
Does Paul have a hedge?
I am somehow surprised at how stats are loosely interpreted to try and derive hedges...
And the game below only involves "mild" randomness!
As a learning trader and investor, it's nice to have someone with experience and access to professional traders sharing his insights and techniques. Thank you!
I have to echo John's concern, though. I'm skeptical of market arguments from inductive reasoning especially when the number of observations doesn't provide much to work with. Rob Hanna over at Quantifiable Edges posts a lot of these research pieces, and I admire his work ethic. But I'd love to know how often these backward-looking tests work out given the ever-changing nature of the market. I'm not a statistician, but can more tools be brought to bear to quantify how much of an edge really exists here, and how many times one would have to make the same trade (e.g. S3/S1 pattern short-term edge) before the likelihood of showing a profit is above X%? How often does a demonstrated edge in the past translate into one going forward?
Post a Comment