From A 50-Day High To A 50-Day Low In 8 Days

Amazingly, Thursday’s close marked a 50-day low in the S&P. It was just last week that the S&P closed at a 50-day high. Moving from a 50-day closing high to a 50-day closing low so quickly is quite rare. I only found 6 other instances. Unfortunately, while it appears rare, it doesn’t appear predictive. Below is the stats table.

Examining the individual charts left me with no deeper insights. I’ve listed the dates of each instance in case anyone else would like to have a closer look.

Poor Closes Going Into A Fed Day

Wednesday is a Fed Day. I’ve written a lot about Fed Days and they’ve historically shown a positive bias. Despite this bias they represent an event that is often anticipated with some anxiety by market participants. This anxiety is natural as participants await potentially market-moving news. What’s interesting is that those times where anxiety is the highest have typically proven much more profitable. To demonstrate this I examined where the SPY closed within its daily range. I used SPY rather than SPX for this test because the daily range is typically more accurate with the ETF thanks to the staggered market opening. Below are all times like Tuesday where the SPY closed in the bottom 25% of its daily range prior to a Fed Day.

Stats here are strongly bullish. Last night’s Subscriber Letter examined the results in even more detail. It also showed what happens when SPY closes in the TOP 25% of its daily range on the day before a Fed Day. You may sign up for a free trial subscription by clicking here if you’d like access to that report.

You may also check the Fed Day label to see previous blog posts about Fed Days.

Poor Breadth On Bounce Somewhat Discouraging

I’ve shown numerous studies over the last couple of years that illustrate weak bounces from oversold conditions are often followed by downside.

A study that appeared in last night’s Quantifinder is an example of this. The study was last shown in the 6/24/09 blog post and is updated below:

Notable about the above study is that 4 of the last 5 instances showed positive returns 5 days out. The one instance that never closed below the entry was the last instance on 11/2/09. Still, the stats are convincing enough that I’m not inclined to completely ignore them.

Gaps Up From 10-day Lows

SPY is set to gap up just over 1% as I type this morning. I’ve shown before that large gaps up from low areas will often spark short covering rallies. Let’s look at some stats. First, here is a look at 1% + gaps up from 10-day lows (all stats look at the last 17 years):



Decidedly bullish edge here. The average loss size suggests risks are high, though.

Buy what if the market pulls back and the gap up is only between 0.5% and 1%?



Now you’re looking at basically a coinflip.

VIX Stretched Over 20% From The 10ma While SPX in Uptrend

The VIX provided some notable action Thursday. It closed up over 19% and is now stretched over 20% above its 10ma. Below is a test of other times we saw the VIX stretched this much while the SPX was trading above its 200ma.



The VIX is one of several indicators I am seeing that is suggesting a bounce is near.

Why I Always Look Deeper Than The Stats Table

The evidence I most often show when illustrating a study is a statistics table like the one below. But it’s not all I look at and it never tells the whole story. In the subscriber letter I’ll often go into more detail on some of the studies. Tonight I thought I’d show an example of one study whose stats table I consider to be a poor representation of the truth.

Tuesday’s rally was the biggest % gain in at least 10 days. It followed Friday’s selloff which was the biggest % drop in at least 10 days. With the market trading above the 200ma and a new 10-day high being made on Tuesday, it made for an unusual setup. Below is a stats table showing similar setups in the past.

From the stats table it appears there is a fairly strong inclination for more upside over the next several days. Now let’s zoom in a bit on some of the results. I chose to zoom in on the 6-day here exit since that had the highest win %. Below is an equity curve with the 6-day exit strategy.

While the surface stats looked good, this chart tells a much different story. For one, there haven’t been any instances in nearly 10 years. Also, the 10 years prior to that there were only 5 instances and the return from them was breakeven. In other words, it’s been over 20 years since any edge has been exhibited by this study. In fact, just about the entire “edge” appears to be thanks to the 80’s. So while the initial results looked substantially bullish, this is definitely not a study that I would want to base a trade on. Traders who conduct their own studies should keep this lesson in mind. It's important to carefully examine all results and not jump to conclusions based of the first set of numbers.

High To Low In 1 Day Revisited

One study that popped up in the Quantifinder Friday night is from the 6/16/09 blog post. It looked at quick moves from a high to a low. I’ve updated that study below:

Such sharp moves from a high level often see spillover in the following days.

2 Up Days In A Market Dominated By Upside Momentum

In August of 2008 I wrote about a simple system that looked to take advantage of the choppy market environment by shorting any time SPX closed up 2 days in a row. I tracked the system and used it as an indicator in the subscriber letter from August of 2008 until August of 2009. In August of 2009 it was becoming apparent that the market had shifted from one dominated by chop to one dominated by upside momentum. In last night’s subscriber letter I ran some studies that showed just how dominant short-term momentum has become. Below is one of those studies:

We see here that since the middle of August, 11 of the 12 times the market has gone up 2 days in a row it has managed to follow through with more gains the next day. It’s also shown a profit 2 days later on 11 of 12 occasions. Shorting 2-day upmoves has been an exceptionally poor strategy in this environment. There are at least a couple of ways traders could use this information. 1) They could look for strategies that try and take advantage of this kind of momentum. 2)They could also simply monitor the action of “2 Up Days” going forward to try and get a feel for when momentum may be waning and the environment changing.

If you’d like to see more on my recent short-term momentum study, you may take a free trial of the Quantifiable Edges Subscriber Letter by clicking here (email address required).

The One Study I Saw With Bullish Implications

Much of what I looked at last night suggested additional downside in the next few days. Below is a copy of the one study that showed short-term bullish tendencies.



{edit - should read highest volume in 5 days}

Declining SPY Volume At New Highs

Over the holidays it was easy to ignore the low volume and dismiss it as typical holiday traffic. Traders may have now returned but trading volume is still lacking. The low volume isn’t blatantly obvious because it’s still higher than it was during the holidays. Still, it’s definitely going in the wrong direction here. Below is a study I looked at last night using SPY.

The edge isn’t huge here but it does suggest a bit of downside is likely in the very near future. (And based on the futures the pullback could be starting this morning.)

Aggregator System for Swing Trading The S&P 500 Finishes 2009 Up Over 36%

A couple of months ago I discussed the Aggregator System on the blog for the 1st time. The Quantifiable Edges Aggregator is a tool I introduced just after I began publishing the Subscriber Letter. The Aggregator tool looks at two things:

1) The net expectations from the active studies published in the Subscriber Letter and on the blog over the next few days.
2) The performance of the S&P versus expectations over the prior few days.

The Aggregator System produces mechanical signals based on the values supplied by the Quantifiable Edges Aggregator. Its signals are based on the S&P 500 cash index and all trades are assumed to take place at the close.

While it can sometimes be difficult to anticipate whether that night’s research will suggest a bullish or bearish tilt and whether it will be enough to shift net expectations, I have been able to incorporate the Intraday Quantifinder into the process to get a fairly accurate read on what the Aggregator is likely to signal that night.

“Probable” signals are published on the Quantifiable Edges systems page about 10-15 minutes before the bell each day. I’ve been publishing them for nearly 4 months now and to this point I believe there has only been one instance where the “probable” signal was switched based on that night’s research.

Once the “probable” signal is posted each day I typically send out notification and a link via Twitter to alert subscribers.

When I first discussed the Aggregator System on the blog in October it was trying to fight out of an 8% drawdown. It did manage to do so and it finished 2009 at a new high. The total reinvested (hypothetical – not including commissions/dividends/interest on cash) return for 2009 was 36.27%. The Quantifiable Edges Subscriber Letter began publishing on 2/25/2008 so that is as far back as Aggregator values go. Total returns for the partial (10+ months) year of 2008 was 60.20%. Therefore the total compounded profits since inception have been 118.31%.

For more general information on the Aggregator System and the Quantifinder you may refer to these posts:

1) The Quantifiable Edges Aggregator
2) The Quantifinder Unveiled
3) Quantifying the Value of Historical Research (The Aggregator System)

A detailed research paper on using the Aggregator System is available on the Quantifiable Edges Systems page in the members’ area. You may also download the complete performance spreadsheet there showing many different statistics and details of all individual trades. All this information may be accessed with a free trial.

I’ve placed a graphic on the right hand side of the blog showing the Aggregator System (hypothetical) equity curve. I will continue to update the performance periodically.

So what does a subscription to the Aggregator System cost? It’s included with all gold memberships. If you’ve never trialed Quantifiable Edges, you may sign up to do so here.
If you have trialed it in the past, but would like another peak before subscribing, just send an email request to support @ quantifiableedges.com (no spaces).

A Study Based On Thursday's Action

Thursday’s price & volume action had a lot of characteristics that would appear bullish. These include the fact that it posted an outside day and closed higher. It also made a new 50-day intraday high. The SPX is in a long-term uptrend and trading above its 200ma. Lastly, the volume rose Thursday as the market rallied. Sometimes it can be interesting to take a number of market observations like these that would seem to obviously suggest a bullish edge and run them through the wayback machine. Below is a study that did just that:



Rather than combining for a bullish edge it appears the scenario above has often been followed by downside.

After An Up Day To Start The Year

One interesting study I showed in last night’s Subscriber Letter was to look at market performance following a positive 1st day of the year. Below are the results:



There appears to be a high probability of immediate follow-through. After the first 1-2 days, though, expectations expectations for the next 2-3 weeks turn negative.

Why The 1st Day of January Has Been So Interesting

Back in July I did a study that looked at the upward bias of the 1st day of each month. As you may recall I ran the study back to 1987 because prior to that there was no significant bias on the 1st day of the month. January is interesting because over the last 23 years its had the worst % profitable of any month but still the 4th highest average returns. When New Years optimism runs high it can create a sharp up move. In fact there have been 3 years (1988, 2002, 2009) since 1987 where the S&P 500 has gained over 3% on the 1st day in January. Below is a copy of the table from the July 1, 2009 blog post with breakdowns by month (stats not updated).