Wednesday, September 20, 2006

The Structure of Market Reversals: What We Can Learn From Yesterday's Market

Yesterday's reversal in the ES market had an almost aesthetic beauty: it so nicely captured the dynamics of market reversals.

What I'd like to propose is that important turns in the market possess a common structure. Once you understand that structure, it's easier to recognize its formation in real time and profit from those reversals.

The chart above was the Market Delta screen I tracked in the afternoon. The first phase of the turnaround is heavy volume at the bid, leading to sharply lower prices. This means that sellers are eager to exit the market--they're not even willing to work orders at the offer to get out--and they outnumber buyers (those willing to take the offer price). Remember that the volume at bid vs. offer is telling us about the very short-term sentiment of market participants. The first phase in the market move above is negative sentiment and downward price movement.

In the second phase, the negative sentiment continues--we still get a preponderance of volume at the market bid--but now there is less downward price movement per unit of downside volume. I used the term "efficiency" in my Psychology of Trading book to describe this relationship between market inputs (volume, for instance) and outputs (price change). The market is becoming less efficient. It is not moving as far in price terms per unit volume as it did earlier. Very often the market makes its ultimate price lows at this phase. Divergences with the NYSE TICK and among sectors are often apparent.

This drop in efficiency precedes major market turnarounds. It can be quantified. Very often, the efficient and inefficient phases are separated by a significant bout of counter-trend activity. We see this in the chart during the 12:00 bar. Buyers took the market higher, with much more lifting of offers than we had seen in prior bars. This tells us that a group of market participants are perceiving value at the new, lower prices.

The third phase is accompanied by significant cross-currents of buying and selling, with the market ultimately unable to print new price lows. During this phase, we typically see many divergences and a positive shift of the distribution of the NYSE TICK. This tells us that, across the universe of NYSE issues, an increased number of stocks are being purchased at their offer price. From the first through third phases, it's not uncommon to see a decline in market volume as selling dries up.

The final phase of the turnaround occurs when selling is exhausted and buyers are emboldened, pushing prices higher on increased volume. Much more volume is transacted at the offer price and now the market gains efficiency to the upside. This upside efficiency will continue until the rise, like the prior decline, faces serious countertrend resistance and begins its own second phase of less efficient, higher prices.

One of the great challenges of trading is recognizing this basic structure across multiple time frames. Note how we made a bottom from July, 2002 to October, 2002 to March, 2003. You'll see a similar process. The recent market bottom in June and July also possessed a similar structure. The longer it takes for the market to go through its phases, the more extended the move in the opposite direction. This, too, can be quantified.

Many of the classic chart patterns (double tops/bottoms, head and shoulders, etc.) are simply price-based depictions of what is occurring in the market auction over time, capturing the movement from phase to phase in market transitions. Pattern recognition is a function of multiple exposure to different varieties of patterns: that's how radiologists learn to read X-rays, for instance. Once you become sensitive to the shifts between efficiency and inefficiency, you'll be able to see patterns set up in real time. I will try to highlight some of these patterns in the Weblog and in my updates.

And that, as in yesterday's trade, can make the difference between profiting from turnarounds and getting run over by them.


James said...


What you've described is essentially a very typical Wyckoff pattern. Of course, wyckoff students might look for several tests on low volume or reactions to the lows on low volume before an up move (as the so called "smart money" pushes price lower to see how much supply is active on lower levels.). On a different note, i would just like to say i think your blog is the best on the net. I check it every day. I just wanted to thank you for sharing.



Brett Steenbarger, Ph.D. said...

Thanks, James, for the kind feedback and the Wyckoff reference. Is there a particular Wyckoff book or info source you'd recommend? I appreciate it--


voodster said...


I am curious about how you refer to market efficiency. Isn't a market that is unable to move much being very efficient? i.e. it is at a value that buyers and sellers are so pleased with that it cannot move much up or down. You use efficiency more to describe a market that is seeking value but finding it easier to travel there. So if it moves several points with less volume it is being quite efficient.

I'm not saying you are wrong or that I stand by my understanding of the term, but could you comment why your interpretation best describes "efficiency"?

Brett Steenbarger, Ph.D. said...


Thanks for your question. I am indeed using efficiency in a process-modeling context, not the usual financial one, so perhaps it's not the best choice of terms. When it takes more inputs to yield a given amount of output, a production system can be said to be losing efficiency. The process modeling tools used in industry to monitor, say, manufacturing processes are relevant to financial systems in that context.