What Are The Chances The Market Gets "Marked Up?"

As we near the end of the quarter I’ve begun to hear quite a bit about the “end-of-quarter markup” phenomenon. I’ve also received a few questions about it. The theory is that mutual funds and other large institutions tend to “mark up” the prices of securities at the end of each quarter so that their return numbers look better. I decided to take a look.

First I ran a test which showed the return of the S&P 500 on the last day of each quarter going back to 1960. Of the 186 quarter-ends over the period, 90 have had a positive last day of quarter, 94 finished negative, and 2 were basically dead even. The average win was 0.065%. The average loss was 0.06%. The net average day was 0.001%. Not even as good as an average day over the period.

I then checked to see what happened if the market sold off the few days leading up to the last day of the quarter (like now). For instance, 7 times the market sold off at least 2.5% in the last 3 days of the month. Five saw gains on the last day and two saw more losses. Unfortunately, the losses nearly eclipsed the gains. Lowering the requirement to a 1.5% selloff in the preceding 3 days gave 18 trades. 9 winners and 9 losers. Net expectancy was slightly negative.

I then looked at what happened if the S&P was down at least 3 days in a row just before the last day of the quarter. Twenty occurrences. 10 winners. 10 losers. Slight negative expectancy.

Looking at recent history rather than all the way back to 1960 did not help these studies.

No matter how I looked I was not able to find any evidence of an end of quarter mark-up in the index. Perhaps mark-ups occur in individual securities, but it’s not apparent in the general market.

Since I figured some people might be getting sick of looking at those “Myth Buster” guys, I posted some other Busters today…

---------------------------------------------------------------------------------------
Would you like to explore these studies more? Or customize them with your own ideas? Quantifiable Edges studies are now available for purchase in Tradestation format along with pre-set workspaces. Click here for more details and a complete list of available studies. Click the buy now button below to download and install directly to Tradestation.

EOQ Markup - $12.00

Review Of Recent Studies

The market has now sold off for two straight days. Of some concern is that I’m not seeing evidence that anything is overdone to the downside. For example, I looked at all stocks in the S&P 100 tonight along with my list of 115 highly liquid ETF’s that I track. None of them made a 10-day low on Thursday. None. Breadth is not suggesting we are due for a bounce.

Price-wise we are back to the middle of the recent range. I have very little to add tonight so I thought I’d do a quick review of outstanding studies. Earlier this week there were several bearish studies which had short-term influence. That influence is beginning to dissipate. Those studies may be found here and here.

Prior to these I had posted several bullish studies with intermediate to long-term influence. Each night in the Subscriber Letter I list all outstanding studies, their time frame and their bias. I find it a useful graphic for helping me organize my thoughts and determine my own trading bias. The bullish intermediate-term studies I consider active are listed below along with the time-frame they looked at.

March 24, 2008 Nasdaq Leadership Bullish - 1-10 weeks
March 19, 2008 Bottom Explosion 2 - 1-20 days
March 19, 2008 3.5% Up Cluster - 10-20 days
March 17, 2008 Consumer Sentiment Stretch - 1-12 months
March 12, 2008 Bottom Explosion - Now What? - 1-20 days

In reviewing them you may notice that many are just beginning to reach their sweet spot.

While the pullback may or may not have farther to fall I am not seeing evidence at this point that it will be anything more than a pullback.

Light Volume Pullback (A Good Thing?)

The standard line on days like Wednesday goes something like this:

The S&P 500 pulled back today on light volume. After the recent run-up the market was due to pull back. The light volume is a sign that selling was not aggressive and should be viewed as a positive. The pullback appears orderly. It would seem buyers just stepped away. There doesn’t appear to be any heavy institutional distribution.

Does any of the above sound familiar? It seems to make sense. Everyone claims they want the market to pull back on light volume. It’s in plenty of books so it must be true. Hmm…

I ran some tests on the S&P 500 looking for the following conditions:

1) Yesterday the 3-period RSI was above 70 (showing there has been a short-term run-up).
2) Today the market closed lower than yesterday.

Buy at the close. Sell “X” days later. Over the last 25 years here is what the S&P 500 has done 3,4, and 5 days out after this setup:


Over the next 3-5 days the market has managed very slight gains.

Next I added a 3rd condition to the mix:

3) Volume must be the lowest volume of the last 10 days.

Again I’m buying at the close and selling “X” days later. Over the last 25 years here is what the S&P 500 has done 3,4, and 5 days out after this setup:


Apparently the “buyers stepped away” on day 1. Over the next few days the sellers filled the void. Very light volume at the beginning of a pullback does NOT appear to be a good thing.

Another myth busted.

More Evidence Suggesting A Short-Term Pullback And Implications If It Doesn't

Last night I showed a couple of studies that suggested the market was likely to begin a pullback or at least a consolidation in the next few days. Tonight I’ll review and remake some past studies.

Tuesday was an inside day for the S&P 500. (Lower high and higher low on the chart.) On February 10th, I discussed inside days with down closes. Tuesday closed higher so it didn’t quite qualify under that study. Looking at all inside days in SPY going back to the beginning of 2001 I uncovered the following:

There have been 215 inside days in SPY since 1/1/2001.
116 times (54%) the market closed LOWER the next day.
The average loss the next day was 0.9%.
The average gain the next day was 0.6%.
The net average move the next day was a 0.2% loss.

I then looked at inside days when the market had made a short-term move up and was at or approaching overbought. For this I required the 3-period RSI to be 70 or greater. This led to the following results:

There have been 48 inside days in SPY since 1/1/2001 with the 3-period RSI closing above 70.
29 times (60%) the market closed LOWER the next day.
The average loss the next day was 0.55%.
The average gain the next day was 0.37%.
The net average move the next day was a 0.2% loss.

The second concept I discussed recently which is once again popping up is consecutive higher closes in a long-term downtrend. Below are the results of selling short the SPY any time it closes higher 3 days in a row while under its 200 day moving average.

More and more evidence is starting to point at a likely pullback. Still, caution is warranted. The market just posted a Follow Through Day. Past Follow Through Days have also typically led to short-term overbought conditions. This did not lead to a downside edge over the short-term. Readers may want to review my Feb. 1st column for more details on this. Also in the Feb. 1st column I show how the first week following a Follow Through Day has predicted the success or failure of the rally about 2/3 of the time. Traders may want to keep this in mind and pay special attention to the action over the next few days.

In short, a pullback now appears more likely than not. Should the market fail to pull back over the next few days that would suggest positive implications for the intermediate-term.

Updated CBI Chart

The Capitulative Breadth Indicator (CBI) returned to “zero” yesterday, finally closing out the last of the recent trade cluster. Below is an updated chart of the index. The “buy” and “sell” arrows on the chart once again show the results if one was to buy the S&P any time the CBI hit 10 and then sell it when it closed at 3 or below. I’ve discussed this crude market timing system in the past. Since 1995 it would now have proved profitable in 18 out of 18 trades – most of which occurred during “scary” selloffs.

For those interested in tracking the trades behind the CBI real time, they are provided to subscribers in the Quantifiable Edges Subscriber Letter. Also in the Letter is CBI percentages of 24 different market sectors.

Market Getting Overdone Short-Term

I looked at a lot of studies in the past week or so and they’ve all said pretty much the same thing: the edge was to the long side. Over the last two trading days the market has shot up impressively. I am now seeing some short-term indications that it is due for a pullback or at least a rest. Let’s look at two quick examples – one price based and one sentiment based.

Price
I looked at shorting the S&P 500 under the following conditions:
1) The S&P closed below its 200-day moving average
2) The S&P rose at least 1.5% the last two days in a row.

Results below ($100,000 per trade):


Based on the price action it appears a pullback is likely to begin soon. A modest edge is apparent to the downside after day 2.

Sentiment
I then looked at the position of the VXO – as a gauge of sentiment rather than price. I looked at the performance of the S&P 500 under the following conditions:
1) The S&P closed below its 200-day moving average
2) The VXO closed at least 10% below its 10-day moving average.

Shorting the market under these conditions and covering when the VXO closed back above its 10-day moving average would have produced the following results since 1987. 51 total trades. 31 (61%) winners. Average winning trade = 2.2%. Average losing trade = 2.6%. Expected value = 0.3% per trade. Profit factor = 1.3.

Not an overwhelming downside edge here, but more evidence that continued upside may not be in the making.

Based on price and sentiment measures, a pullback beginning in the next few days seems to be a likely scenario.

Nasdaq Taking Leadership Helps Bullish Case

A few weeks ago I posted a study which discussed the implications of the lagging relative strength of the Nasdaq versus the NYSE. The results of that study were strongly negative from 1-10 weeks out. Last week I noticed the Nasdaq was trying to take the lead back from the NYSE. When the market closed for the week the Nasdaq did manage to barely overtake the NYSE based on the indicator I’ve described in the past. A very astute subscriber to the Quantifiable Edges Subscriber Letter also noticed this and sent me a note asking me to comment on the significance in light of my study of a few weeks ago. I’ve updated the chart from a few weeks ago below. The red line has now barely crossed the yellow line signifying the Nasdaq’s RS is slightly stronger.


I ran a few studies to determine the possible significance of the Nasdaq taking the lead. The first one simply looked to buy any time the Nasdaq went from “lagging” to “leading”. On the chart above this would be shown by the red line crossing above the yellow line. The results are below.


Overall the results were generally positive going forward. I then looked at situations like the present where the Nasdaq rebounded from an extreme lagging position as described in the previous study and then crossed over to take the lead. There were 15 such occurrences. The performance of the NYSE from 1-10 weeks out can be found in the table below.

It appears this situation is more even more favorable than a typical cross.

I then adjusted the exit criteria to change this study into more of a system. Rather than exiting “X” weeks later, I said to exit whenever the red line re-crossed below the yellow, which would represent the Nasdaq falling back to a lagging position.

Buying when the RS upward cross occurred and selling when the Nasdaq’s RS line crossed back below the NYSE would have produced 11 winning trades and 4 losers. The average winning trade was good for more than 5% and the average loser saw a decline of less than 1.2%. The length of the average winner was 7 weeks vs. 5 weeks for a loser. If $100,000 was allocated to each trade (assuming no commissions or slippage) the gross profits on winning trades would have been $50,901.45 versus gross losses of $4,617.18 for the losing trades. This equates to an outstanding profit factor (gross gains / gross losses) of 12.02.

It appears the change in the Nasdaq from “lagging” to “leading” status is another argument for the bullish case.