Where The Edge Lies When The Market Drops X Days In A Row

A few weeks ago I wrote a post that illustrated the edge buying 3 down closes has produced over time. There are a lot of ways to expand on the research in that post. Today I’m going to look briefly at what its meant to be down “X” number of days in terms of an edge for the next trading day.

The table below looks at performance the day following a drop in the SPX of exactly X days in a row.

A few things caught my eye here. The first is the % Wins column. It stayed pretty close to 50% no matter how many down days you’ve had. Basically a coin flip either way. The 2nd column that caught my eye was the Avg Win column. This is the column that provides an edge when you get further out. In general the more stretched market the stronger the bounce is likely to be.

As I showed in the post that focused on 3 days down, the edge has changed over the years. More recently there has been a greater tendency for the market to reverse rather than trend. Below is the same table for the 1989 – present time frame.

The % wins column here shows that the tendency to reverse after being down several days has grown stronger in the last 20 years. As in the 1st table, the Avg Win increases as the SPX gets more stretched.

I demonstrated with the “Down 3 Day” trade that much of the edge has come quite recently. Below is an equity curve for being down 5 days.


Much the same here. A good majority of the edge is attributable to the last several years.

My takeaway is this. There is a long-side edge after the market has pulled back for several days in a row. A large part of the edge comes not from the chances the market will rise the next day, but the fact that you will be rewarded amply if it does. If you understand this and perhaps combine the fact that the market is down multiple days with other indications that a reversal is likely, then you can likely identify a strong upside edge.

Nasdaq/NYSE Volume Ratio Hitting Extreme Levels

The Nasdaq/NYSE Volume Ratio is an indicator I haven’t discussed on the blog, and not too often in the Subscriber Letter. It is hitting extreme levels at this time and deserves some attention. One word of caution - levels will vary depending on data provider. So while the extremes may differ depending on whose data you use, results should be comparable at those extremes. I use Tradestation. On Tuesday the Nasdaq/NYSE volume 20-day average closed over 1.65. Below is a table showing 1-month returns based on this ratio.

Nasdaq / NYSE 20-day volume ratio exceeds X. Buy S&P 500 on close. Sell 20-days later. $100k/trade. (click to enlarge)

High levels of Nasdaq trading as opposed to NYSE suggest excessive speculation by investors. Once this level exceeds 1.4 (as Tradestation measures it), is has generally indicated a bearish bias.

Top Weighted Nasdaq 100 Components Very Overbought

I noted in Thursday night’s Subscriber Letter that AAPL and GOOG had both risen for 8 days in a row. Additionally, those two along with MSFT and ORCL all had 2-day RSI’s of over 98. That’s an extremely overbought level. These 4 stocks are among the top 7 and make up about 26% of the Nasdaq 100. Using the list of current Nasdaq 100 stocks I studied action among the current 8 highest weighted. In addition to the 4 above this includes RIMM, QCOM, CSCO and GILD. I looked at other times since 2007 that at least 4 of these 8 stocks closed with a 2-day RSI in excess of 94. Those results are below:



It’s not exactly a layup that a pullback should immediately begin. Still, risk appears to greatly outweigh reward when several of the top components are strongly overbought short-term. A brief look at the W/L Ratio suggests this. A pullback does normally come at some point in the next few days though. In fact of the 21 instances where the conditions were met, only one did not experience a close below the trigger day’s close within the next 4 days.

Nasdaq New High & Low Volume Spyx Suggest Edge

One notable statistic from Thursday’s action was the Quantifiable Edges Nasdaq Volume Spyx indicator closed around minus 4. (For those uninitiated my volume spyx indicators look at comparative volume across multiple securities. When ratios get out of whack, it often shows up as an upside or downside spike on the chart.) Closes below zero are rare and often lead to weakness over the next few days. This is especially so when the market rises along with the low readings. Below is a study that exemplifies this.


The number of occurrences is a bit low but certainly suggestive of a downside edge over the next several days. The edge appears the strongest over the 1st 2 days, when much of the damage has been done. Not evident above is that 13 of 14 instances closed lower than the trigger-day close at some point in the next 3 days.

In order to gain a larger sample size I also looked at Nasdaq Spyx readings below 10.


Results here are similar to the 1st test, but with a decent sample size. This all suggests a downside edge in the Nasdaq 100 over the next few days.

Clusters of large low-volume selloffs

Last August I looked at clusters of large low-volume selloffs. While the S&P hasn’t quite met the parameters laid out in that blog post, it has met the loosened requirements which were looked at in the 8/1/08 Subscriber Letter. The Quantifinder picked up on this last night and I thought it was worth another look. The basic concept is as follows. Strong moves down often lead to bounces. When you have a series of them occur on low volume then you may have favorable risk/reward on the long side.





There appeared to be an upside edge when I ran this test last summer and there still appears to be one.

Tweaking The Nasdaq/S&P Lead/Lag Model

Last week I discussed an indicator designed by Gerald Appel and published in his book “Technical Analysis – Power Tools for Active Investors”. He refers to it as the Nasdaq/NYSE Relative Strength Indicator (not to be confused with RSI). I changed it slightly for my purposes and instead of tracking the Nasdaq vs. the NYSE, I instead tracked the Nasdaq vs. the S&P 500. In last week’s post I demonstrated that the S&P had performed much better over time when the Nasdaq was in a leading position.

I also provided a spreadsheet with the calculation and a model based on the indicator on the free downloads section of the website.

When conducting research or designing models it’s important to avoid just looking at it from one angle. Today I’m going to explore a couple of other ideas that the indicator / model may have invoked in many of you. I’ve also updated the spreadsheet so that you can see how these ideas were tested as well.

First, a quick refresher chart of last week’s results.

As you can see, utilizing the indicator would have greatly enhanced your returns over time. So if owning the S&P when the Nasdaq is leading is so favorable, does that mean you should short the S&P when the Nasdaq is lagging?

Below is an equity curve of a $100,000 investment doing exactly that. (Interest and dividends are not included.)

After almost 40 years the current (recent?) bear market just got you back into the black. Certainly this isn’t an equity curve that suggests an edge. While some may be surprised based on how positive the previous results were, it makes complete sense. The strategy in this chart calls for shorting the leading index. I don’t believe I’ve ever seen anyone who has suggested that to be a good idea.

But what if instead of buying the S&P when the Nasdaq leads, we buy the Nasdaq? It is the leader after all.

Equity chart below. Hold on to your hats.


Like the original test, dividends are not included. The annual growth rate if you earn 2.5% interest on your cash balance is about 13%. This is more than twice the S&P 500 annual growth rate of under 6% for the period. There’s also lower drawdown and you’d only be in the market a little more than half the time.

Last week we saw how timing the market with this simple indicator can make a big difference. Here you see that a small tweak to ensure you’re in the leading index can juice returns much, much more. There are numerous other tweaks you could add to the model to improve performance or reduce drawdowns further. This is as far as I’m going to take it, though. The spreadsheet is still available and updated with the above tests and charts. You may download the updated version on the free downloads page. I’d suggest anyone who downloads it should use it as a starting point – not a finished product.

For more ideas on using Excel for historical analysis I’d recommend buying Dr. Steenbarger’s Daily Trading Coach book. A few weeks ago I created a sample spreadsheet based on the lessons in that book. You may download that spreadsheet on the free download page as well. Registration is required for the Nasdaq/S&P Lead/Lag Model. Registration is NOT required for the Daily Trading Coach Spreadsheet.

An Incorrect Assumption

6/3/09 edit CXO has now deleted their old post and re-run the tests using the correct indicator. Their corrected post can be seen by using the link below.

CXO Advisory posted a column this morning in which they attempted to refute the effectiveness of the 10-week Nasdaq/S&P 500 lead/lag indicator. In their column they referred to my post of last week and Gerald Appel’s book which I referenced as the origin of the indicator. They then ran several unrelated tests to show that 10-week RSIs are not effective in determining future returns. It is important to understand that RSI was not the indicator described by Gerald Appel nor used in my testing.

Frankly, in the past I’ve found some interesting things on their site, but their lack of attention to detail here is flabbergasting. If they had either 1) opened the book, or 2) downloaded the spreadsheet I provided for FREE with the complete research supplied, they would have understood this. Instead they ran their own tests using a completely unrelated indicator.

Their tests ran from 2005-present. They showed that over the 2005-present time period buying the S&P when the 10-week RSI of the Nasdaq was above the 10-week RSI of the S&P would have lost money. This doesn’t surprise me.

For anyone who has downloaded the free spreadsheet, if you plug a round number (such as $1,000 or $100,000) into the 12/31/04 row you can see the following results from then until 5/22/09. The S&P would have lost about 25% (not including dividends). The model (assuming 0% interest on cash and not including dividends) would have gained about 11.6%.

In the next few days/weeks I will be publishing some more research that pertains to this indicator. I hope readers will find it valuable.

In the future should CXO attempt to refute the work of others I would hope they at least make an effort to look at the work they are refuting.

One final note, the Nasdaq/S&P 500 lead/lag indicator signaled a buy on Friday’s close. Subscribers to Quantfiable Edges are now notified of all model changes via the new Quantifinder technology.