sentimenTrader Blog


2016-01-07 | Jason Goepfert | Comments

In terms of panicky opens, we’re now in rare territory.

Since the inception of the S&P 500 futures market in 1982, we have rarely seen such a concentrated period of traders selling so heavily before the open of regular trading. Typically this is seen during instances of worries about market structure (e.g. the flash crash in 2010, Black Monday in 1987) or global financial meltdown.

Barring a large recovery into the open, this will be the third opening gap of -1.5% or more in four days. The only times we’ve seen this kind of opening pressure were during the 1987 crash, the peak of the financial crisis in 2008 and during the flash crash of 2010.

20160107_gap

If we just look at gaps of -1.5% or more that occurred at a three-month (but not one-year) low, then we get the following. The only other time we’ve seen this on back-to-back days or close to it was during the 2010 flash crash.

20160107_gap2

We can also look at times when there were back-to-back gaps of -1.5% or more, as shown below. Mostly the same time periods.

20160107_gap3

Shorter-term, this is more equivalent to panic conditions that trigger impulse buying and form shorter-term lows, lasting days to weeks. It is not typically the kind of behavior seen at medium-term lows of several months or more.

Most often after periods of extreme volatility like this, there is a period of testing of the panic lows that goes on for months, and which are more likely to generate the kind of pessimistic sentiment readings that we’re not seeing yet, because most of the selling pressure has been focused at the open.


2016-01-06 | Jason Goepfert | Comments

Futures are indicated to gap down heavily again this morning, with the S&P 500 currently looking at an open of -1.8%.

Barring a recovery before the open, this would be the second gap open of -1.5% or more in the past three sessions, an event that last occurred nearly five years ago.

The table below shows every time since the inception of the futures in 1982 when there was a second gap of -1.5% or more within three trading days.

20160106_gap

Returns going forward were highly positive but also volatile. When stocks rallied, they rallied big; when they declined they declined big. This isn’t a surprise, but six months and one year later, returns were extremely positive, double an average return, even including the negatives from 2008 which made up the bulk of the instances.

What’s unusual about our current occurrence is that it is happening so close to a recent high. Today’s open is indicated to be about 7.5% from the highest close of the past year. This would be the smallest the S&P had ever declined from a peak before suffering such gaps.

The only other instance when the S&P was within 10% of a 52-week high was October 1997, after which the futures rallied almost right from the get-go.

The only others when within 15% of a 52-week high were in May 2010, after which there was some back-and-forth but again mostly limited downside in the weeks ahead.

There were only two others that occurred in January (2008 and 2009). Even during the teeth of the bear market, these gaps led to choppy rebounds. Granted, that was from what had been sell-offs much heavier than what we’ve seen this time.

Opening gaps like this are disconcerting, especially given some of the concerns we’ve outlined recently in the Daily Reports, but if we accept that opening prices are mostly “dumb money”, reactionary trading, then we should interpret these gaps as temporary and likely to lead to higher returns in the weeks ahead.


2016-01-05 | Jason Goepfert | Comments

We’ve discussed the Turnaround Tuesday pattern many times over the past 15 years, and it’s something that continues to hold up.

Basically the idea is that if the market trends strongly over the few days into Tuesday, then we often see a reversal during mid-week.

Since tech stocks got hit particularly hard, let’s look at every time QQQ lost at least 1% on Monday then gapped down on Tuesday morning (Figure 1). Futures are recovering at the moment but still showing a slight loss. We’ll restrict the instances to those when QQQ was trading above its 200-day average.

20160105_qqq

From Tuesday’s open through Thursday’s close, QQQ rallied 70% of the time with a positively skewed risk/reward ratio. There were three losses of more than -2% (including a doozy in April 2000) compared to ten winners of more than +2%.

If we stipulate that Monday’s loss was -1.5% or more, then we get 9 winners out of 11 occurrences. That huge loss from 2000 is still among them, though the other loser was only -0.2%.

Combined with a few severely oversold short-term indicators (such as the Down Pressure readings over the past three days) and Monday’s late-day reversal, the Turnaround Tuesday pattern has some potential. As noted in the Daily Report yesterday, sentiment still is not showing any extreme pessimism overall, so confidence is not particularly high.


2016-01-04 | Jason Goepfert | Comments

With the day winding down, there was no sustained bounce in the major indexes or most sectors that were hit hard at the open.

This kind of activity tends to impact sentiment, so below is a preliminary look at today’s Optimism Index readings for the most active ETFs that we track.

20160104_etfs

Consumer Discretionary and Financials are near the bottom. With these raw daily figures, we typically look for something below 5 or above 95 to indicate a true daily extreme (using a moving average, the extreme thresholds are more like below 20 or above 80 for most of the funds).

Among the funds currently showing the most pessimism, the one with the most consistent tendency to snap back was Financials, XLF. The table below shows the returns if buying $100,000 of XLF on any day that its Optix closed at or below 5 and holding for the given number of days. It tended to be a short-term phenomenon only, mostly due to a number of days during 2008.

20160104_etfs_xlf


2016-01-04 | Jason Goepfert | Comments

It’s ugly out there this morning with almost anything equity-related indicated to open down 1% or more.

Patterns and stats are likely useless at the moment. Traders and investors are going to be scrambling to adjust after a rough year for many in 2015. Nobody wants to start the year with a big loss that’s going to take weeks to recover.

About the only safe bet is that volume is going to be high.

There is no precedent for such a large gap down to open a new year, at least not since index futures were introduced in 1982. The previous largest gap down open on the first day was in 1984 when the futures opened down -0.8%. Stocks rallied right from the get-go that day and tacked on about 3% over the next few days.

For what it’s worth, if the S&P loses 1% or more today, here are the other times it has done so since 1928 on the first day of the year:

20160104_firstday

It rallied over the next month 9 out of 12 times, averaging excellent gains though the losers were all large as well.

As discussed on Friday, the overall environment is questionable at best, and there are few indications at the moment that trying to buy into the gap down is a low-risk, high-reward scenario.


2015-12-31 | Jason Goepfert | Comments

Shocking events have a tendency to mark sentiment extremes.

We see it time and again across markets. Any development that forces traders and investors to instantly and substantially alter their outlooks has a tendency to create a market move several standard deviations from the norm. Markets then spend weeks, months or years recovering from the shock as more information becomes available and cooler heads prevail.

On December 17, the new political administration in Argentina lifted some currency controls. While it may have been expected, it triggered a massive revaluation of the Argentine peso.

We wouldn’t pretend to be knowledgeable about the fundamental impact of the currency valuation on Argentina’s economy, currency flows or political popularity. Our focus is on sentiment, and what previous devaluations have triggered.

Our friend Steve Sjuggerud recently highlighted the idea that this is likely a good thing. And, indeed, it usually has been, especially longer-term.

The chart below shows the Merval Index versus the spot value of the peso. All monthly moves of more than +5% in the ratio of the dollar to the peso are highlighted with an arrow (if the blue line is rising, then it means it takes more pesos to buy one dollar, i.e. the peso is declining in value). 20151231_peso_chart

Figure 2 shows the returns in the Merval index after any monthly move of more than +5% in the dollar/peso ratio. As we can see, it led to a higher value in Argentine stocks a year later every time and by an astounding margin.

20151231_peso_table


2015-12-30 | Jason Goepfert | Comments

When people discuss seasonality, it is usually in terms of how often, and by how much, the general stock market rallies or declines during certain times of the year.

Those effects have mostly deteriorated over the years as they have become more well-known, but there are still pockets of consistency the further away we get from the broad stock market, such as in commodities or in individual stocks and sectors.

Alpha Architect recently highlighted a piece of academic research suggesting that seasonality was strongest in December and January for momentum names. Those with positive momentum did best in December, and those with the worst momentum did best in January. Investors bought winners at the end of the year and sold the losers to harvest tax benefits, then reversed it in the New Year.

Looking at our list of the most active ETFs for which we track sentiment, here are the best and worst performers so far in 2015:

20151230_seasonality

Over the past 15 years, the top-performing ETFs returned an average of +0.2% during the first week in January and were positive 53% of the time. The worst performers averaged +1.7% and were positive 73% of the time, so there was some evidence of mean-reversion on this level.

The effect dissipated the further out we look, as we’ll discuss in the Daily Sentiment Report this evening. But for those looking to get in (or out) of shorter-term trades in the coming days, it pays to be aware of this seasonal mean-reversion effect.


2015-12-28 | Jason Goepfert | Comments

A feature of our charts that some may not be aware of is using moving averages alone on a chart, to take a longer-term view and get rid of some of the noise.

Once you pull up a chart, select a moving average from the drop-down box. Then click the indicator name in the legend (which makes everything disappear), then click the name of the moving average in the legend. That will plot only the moving average on the chart, which helps clean it up.


2015-12-28 | Jason Goepfert | Comments

A Bloomberg article today noted an outflow from commodity ETFs this year, an exception to past years.

It’s an illuminating item, however it does get one thing wrong, which is something we will commonly see during year-end reviews.

It noted that energy ETFs saw an inflow of cash as investors tried to bet on a bottom.

Source: Bloomberg

As we’ve discussed multiple times, however, this is misleading. Many times, commodity funds, in particular, have seen a sudden “infux” of money as shares are creating while the funds are diving lower.

But this is not driven by investors betting on a rebound. Many of the new shares are created to satisfy a surge in demand for short sales. So, in fact, what looks like an inflow is actually an outflow. Witness 2008 in the chart above.

This is an important point we need to consider when looking at ETF flows. If the flow is going against the immediate trend, then it is likely not a contrary indicator.


2015-12-24 | Jason Goepfert | Comments

We all know that stocks, in general, tend to rise during the last week(s) of a year through the first few sessions of the New Year. But what about other markets?

Seasonality is a tricky thing. It’s easy to see past performance, but there is some game theory involved as well. If everyone knows something, then the time frames can shift as traders try to front-run the masses.

Instead of looking at the broad market, let’s check the Most Active ETF list and see how they have held up during this part of the year. The table below lists them from best- to worst-performing, assuming we invested $100,000 in each trade.

20151224_santa

Surprisingly, commodity and basic material ETFs were at the top of the list of most consistent winners. In the silver ETF, we would have generated a profit of over $33,000 despite being in the market for only 82 trading days.

The most consistent losers were volatility (no surprise there), along with the most interest-rate sensitive funds like TLT, VNQ (REITs) and XLU (Utilities).

Stocks have already rallied heartily heading into the season, perhaps evidence of the anticipation effect noted earlier. Assuming the pattern holds up, though, we should see rate-sensitive issues under-perform to the benefit of what had been some of the most beaten-down funds of the past year.


← previous page · next page →