Tuesday, September 29, 2009

Low Volume When The Market Rises Strongly

I showed yesterday how a very-low volume day during a decline can often lead to a short-term reversal. Today I will review a study that first appeared in the blog on 5/13/2008. It looks at extremely low volume on strong up days - like Monday. (Volume studies typically use the symbol $TVOL in Tradestation, which is their measure of NYSE volume. This is what is being used in the below study.)

We’ve seen several studies like this over time and many of them were identified by the Quantifinder on Monday. With so many studies confirming each other, it seems the downside edge in these very low volume rises is for real. One caveat with Monday’s action though is that it was Yom Kippur, meaning a lot of traders were out of action and somewhat lower volume could be expected. Still, it’s been a steady enough edge that I decided to it was worth review.

Monday, September 28, 2009

Friday's Very Low Volume Provides An Upside Edge

One hint from Friday that suggests a bounce is likely is the fact that volume came in very low. It was the lowest volume in over 2 weeks. Below are some tests which demonstrate the potential importance of this.

First, let’s look at 3 day pullbacks that don’t occur on extremely low volume.
(click table to enlarge)

As you can see there is a slight upside edge. The % winners were between 57% and 60% and the average gains days 1-3 were between 0.1% and 0.275%.

Now let’s look at times were the volume was extremely low:
(click table to enlarge)

Here the upside edge is significantly stronger over the next few days. The % winners were between 68% and 76% and the average gains days 1-3 were between 0.59% and 0.94%. The reliability of a bounce occurring at some point in the 3 day period increased from 75% to 87%.

Volume can often provide some valuable clues. For more volume-related analysis, you can check out the volume label on the right hand side of the blog.

Thursday, September 24, 2009

Fed Day Selloffs

Wednesday’s Fed day reversal acted much like the one that appeared on April 30, 2008. The May 1, 2008 blog featured a couple of studies that were relevant again today. They were just some of what the Quantifinder identified as relevant last night. I re-ran those studies below. This 1st one looks at times the SPX made a 20-day high and then closed in the bottom 10% of its daily range.
(click table to enlarge)

There appears to be an upside edge here. Although it isn’t the most powerful edge we’ve seen it still appears to suggest bullish inclinations over the next few days and weeks.

Also in that May 1, 2008 blog I looked specifically at Fed days that closed in the bottom 10% of their daily range. I re-ran that study as well tonight.
(click table to enlarge)

Much of the edge here appears in the first day. Poor closes on Fed days have rarely seen significant follow through in the coming days. More often the selling is viewed as an overreaction and the market is able to rebound a bit.
It appears Wednesday's action suggests bullish implications over the short-term.

If you’d like to trial Quantifiable Edges premium services including the Quantifinder and the Subscriber Letter you may sign up for a free trial here. (Those that have trialed before but not since the Quantifnder was released in June may email support @ quantifiableedges.com (no spaces) and I’ll be happy to set you up with a new trial.)

Wednesday, September 23, 2009

A Long-Term Look At Fed Days

Below is a long term chart of market performance on scheduled Fed days. I didn’t include non-scheduled meetings. Those are generally surprise rate cuts that are aimed at boosting the market. They’re inherently bullish yet unpredictable since the meetings aren’t scheduled. Therefore there is no point in including them in this study.
(click chart to enlarge)

Over the last 27 years there’s been a persistent upside edge. The average Fed day has ourperformed the average day by about 7.5 times.

For more studies on Fed days make sure to check out the Fed day link.

Monday, September 21, 2009

Last Week's Equity Put/Call Ratios Suggest A Pullback

The CBOE Equity put/call ratio was consistently low last week. The 5-day average is 0.53, which is nearly 22% below the 200-day average of 0.68. When the 5-day average gets extremely low as it is now that can lead to a short-term market pullback. Below is a study that exemplifies this.

Most notable and also most compelling about this study is the fact that all instances have occurred after the March bottom. Prior to that the 5-day ratio had never stretched 20% below the 200-day. This is another example of just how extreme the current bull move has become. It also makes the results that much more impressive from a negative standpoint since they were achieved during a 60% run-up in the market.

Thursday, September 17, 2009

The 2009 Rally - Breadth Without Compare

Yesterday I looked at Worden Bros. T21111, which measures the number of stocks trading at least 2 standard deviations above their 200ma. As you’ll recall, it was hitting an all-time high. (Data goes back to 1986.)

With Wednesday’s big rally, we are now seeing even more extraordinary numbers. Not only is T2111 up to 58.51%, but T2112, which measures the % of stocks trade at least 2 standard deviation above their 40-DAY moving average, is also in record territory. It is showing that a remarkable 57.19% of stocks are now stretched far above their 40-ma’s.

The action in T2112 truly exemplifies the uniqueness of the rally since March. Below is a long-term look at the indicator. Note that from 1986 through 2008 the highest reading this indicator ever registered was 37.22% in November of 2004. That record has been blown away repeatedly over the last 6 months.

Let’s now zoom in on this year to better see what I’m saying.

There simply is no comparison over the last 23 years to what we are seeing in this recent rally. There have now 5 distinct periods in the last 6 months where T2112 has rallied through the old high. And now we’re seeing the most extreme breadth numbers of all.

Wednesday, September 16, 2009

Never Have So Many Stocks Been So Stretched Above Their 200ma.

Near the end of August I discussed that some of the breadth measures tracked by Worden were near all-time highs. This situation corrected itself as the market embarked on a brief selloff. Tonight two of their indicators actually registered their highest readings ever. These are T2109 and T21111 which track the number of stocks 1 and 2 standard deviations above their 200-day moving averages. Below is a long-term chart of T21111 with full history of the indicator going back to 1986.

I marked on the chart the 4 other instances that came close to the current reading. What you may notice is that these spikes were generally brief. Every case was followed by at least a mild selloff that worked off the severely overbought conditions. In no case did the extreme spike mark the end to the bull market that created it. It’s dangerous to read too much into only 4 instances, but a short-term pullback does seem reasonable. The current reading does not suggest a long-term top, though.

Tuesday, September 15, 2009

The 1st Profitable Close Exit Strategy - When It's Appropriate

Today I just want to touch briefly on the exit parameters for the “2 Days In Chop” systems that I discussed yesterday. The exit strategy is basically a time stop married with a “first profitable close” exit. For many traders, a “first profitable close” exit may seem like nothing more than a ploy to inflate the winning % of the system and not an appropriate exit technique to put into practice.

Sometimes this is true. In other cases though, the 1st Profitable Close exit is appropriate and effective. “2 Days In Chop” is one of those cases. Recall the premise of the system was based on taking advantage of the extremely choppy market conditions that had been identified. It’s those choppy conditions that make the 1st Profitable Close strategy viable.

When conditions are especially choppy and the market is constantly swinging back and forth, the expectation is for that chop to persist. This would suggest a move in the direction of the trade is more likely to be reversed than to follow through. So with a system like “2 Days in Chop”, the expectation flips as soon as the trade becomes profitable. Since a reversal is more likely than continued follow through the correct play is to take the quick profit.

Obviously an exit strategy like this only works well when trading a reversal / mean reverting system in a choppy environment. In a different environment, or if trading a breakout system, an exit strategy that looks for quick profits would be a disaster.

Other techniques that work well when trading overbought/oversold conditions would include using a short term oscillator and waiting for that oscillator to revert back to a neutral state, or using a short-term moving average (such as a 5-day) and then exiting the trade on a cross of the moving average.

Monday, September 14, 2009

2 Days In Chop Systems - 1 Year Later

About a year ago I showed 2 systems that looked to take advantage of the market’s choppy nature. Since that time I have tracked the performance of these two incredibly simple systems in the Quantifiable Edges Subscriber Letter. I’ve referred to them as the “2 Days In Chop” systems. As a quick refresher the rules for each are below:

Long System (2 Days Down In Chop):
1) Buy the SPX any time it closes lower 2 days in a row.
2) Sell the 1st profitable close up to 3 days later.
3) Sell on the 3rd day regardless of profitability.

Short System (2 Days Up In Chop):
1) Short the SPX any time it closes higher 2 days in a row.
2) Cover the 1st profitable close up to 4 days later.
3) Cover on the 4th day regardless of profitability.

A few quick notes:
At the time the market was locked in a downtrend which is why I gave the shorts an extra day.
I noted the systems were very raw and were not something I would trade “as is”.

Below I will show the combined performance of the 2 systems since I introduced them. In upcoming posts I’ll discuss how I use the systems and also discuss some thoughts on them and some ideas in which the basic systems could be improved.

Here is the performance over the last year +.

A 73% return would seem very impressive for something so simple. It has had a bit of a drawdown lately, though. Below is a profit curve.

The system peaked on 6/22 and has had a few rough trades as of late. Still, the recent drawdown is very small compared to the overall gains of the system.

So why haven’t I simply traded the system “as is” for the last year? I guess you could say that I’m just not smart enough to blindly trade a system this dumb.

I’m getting a little smarter, though. And I’ll have more in upcoming posts.

Wednesday, September 9, 2009

SPY Rising While SPY Volume Declines

In May of 2008 I showed how 3 higher closes in the SPX (while under to 200ma) had different implications depending on the volume pattern. Tuesday we had this 3-higher closes pattern appear in the SPY while SPY volume declined all three days. This is a slightly different twist and one worth examining:

(click table to enlarge)

This setup appears especially bearish over the 1st week. Possible bearish implications extend out much further than just a week, though.

Would you like to be made aware any time this setup triggers in the future? The Quantifinder does it for you! Versions of the Quantifinder are available with both gold and silver subscriptions.
As I ready to publish this I see that Cobra also noticed this pattern last night. Check out his take as he identifies several recent occurrences on his chart.

Tuesday, September 8, 2009

Labor Day Week Edges

From a seasonality standpoint, Labor Day week has historically been a bit weak.

(click table to enlarge)

Tuesday through Thursday have shown risk/reward to favor the bears, though whether the market is up or down has basically been a coin toss. If we look at times like the present where the market has made gains in the weeks leading up to Labor Day, you’ll find the implication is a bit more bearish.
(click table to enlarge)

This isn’t the most compelling edge I’ve ever published, but for a study based primarily on seasonality, it’s not bad.

Sunday, September 6, 2009

Dog Days of Summer Promo Ending

Labor Day is the last day to take advantage of the Dog Days of Summer promo. Sign up for an annual subscroption to Quantifiable Edges Gold, Silver, or Triple Play packages and receive the Q1 2008 Studies Package ($195 value) free! Offer ends 9/7/09 at Midnight.

More details may be found here.

Friday, September 4, 2009

Is Thursday's Low Volume Troubling?

In last night’s Subscriber Letter I examined whether the relatively low volume on Thursday’s bounce should be concerning. I examined it a few different ways. Below is one series of tests I showed. First let’s look at what happened when volume came in higher on a bounce following some severe short-term oversold conditions.

(click table to enlarge)

Interesting that the high volume bounces failed to follow through. The number of instances is a bit low but the numbers are fairly compelling anyway. Let’s see how this compares to low volume bounces:

(click table to enlarge)

Here the results have a solid bullish tilt.

These tests suggest that Thursday’s light volume should not be of any concern. In fact, it may be a market positive.

Thursday, September 3, 2009

When Months Start Bad Revisited

Beginning of months is a seasonally strong time. The market rises on most occasions. It is fairly rare to see a month start off with two down days. I last looked at bad starts to the month in the blog on June 4, 2008. I’ve updated that study below:
(click to enlarge)

16 for 16 bouncing at some point in the next week suggests a pretty good chance we see a bounce here soon.