ONE-TWENTY TWO - A collection of personal articles on financial markets including analysis you can use
Sep
16

T2108 Update (September 17, 2014) – It Just Took One Man to Light Up the Market Tinder?!

written by Dr. Duru
Bookmark and Share

(T2108 measures the percentage of stocks trading above their respective 40-day moving averages [DMAs]. It helps to identify extremes in market sentiment that are likely to reverse. To learn more about it, see my T2108 Resource Page. You can follow real-time T2108 commentary on twitter using the #T2108 hashtag. T2108-related trades and other trades are posted on twitter using the #120trade hashtag)

T2108 Status: 39.2%
VIX Status: 12.7%
General (Short-term) Trading Call: Hold (TRADERS BE READY FOR WIDE SWINGS)
Active T2108 periods: Day #310 over 20% (includes day #280 at 20.01%), Day #26 over 30%, Day #3 under 40% (underperiod), Day #6 under 50%, Day #8 under 60%, Day #50 under 70%

Reference Charts (click for view of last 6 months from Stockcharts.com):
S&P 500 or SPY
SDS (ProShares UltraShort S&P500)
U.S. Dollar Index (volatility index)
EEM (iShares MSCI Emerging Markets)
VIX (volatility index)
VXX (iPath S&P 500 VIX Short-Term Futures ETN)
EWG (iShares MSCI Germany Index Fund)
CAT (Caterpillar).

Commentary


This is what I tweeted after I blinked and noticed that buyers had already taken the market upward off the lows and to flatline. Little did I know at the time just how true these words would ring.

I came into the day with the T2108 Trading Model on my mind. It was predicting a down day so the sudden gap down seemed to confirm another day of selling. It was not until the market exploded higher that I realized I needed to switch to a different mental model: a quasi-oversold market just looking for an excuse to buy. That excuse came in the form of WSJ journalist Jon Hilsenrath – someone whose impact my poor little model simply cannot anticipate.

Hilsenrath is the WSJ’s “beat” reporter for the Federal Reserve. He is the one who typically gets to attend the press conferences and ask questions on behalf of the WSJ. Over time, he has apparently developed quite a trusted reputation as a Fed-watcher, but I do not think I have ever seen him (presumably) move markets like he did today – stocks and currencies combined!

Hilsenrath had a simple message, a message that I have essentially made over and over myself: “do not overthink the Fed.” He basically argued that the Fed is in no rush to get hawkish on monetary policy and will look to qualify its “considerable time” framework for raising rates such that the market does not pin the Fed down on a specific calendar date. That was enough to send the market higher in sweet relief. The first chart below shows the action on the S&P 500 (SPY) in 5 minute chunks so the changes in trading are more obvious.


In a blink sellers lost control of the open...and then Hilsenrath sealed the deal

In a blink sellers lost control of the open…and then Hilsenrath sealed the deal

Not quite 2000, but a bullish push for the S&P 500 above 50DMA support and after a "hammer" day on Monday

Not quite 2000, but a bullish push for the S&P 500 above 50DMA support and after a “hammer” day on Monday


The NASDAQ also benefited, but it is still underneath a bearish topping pattern.


The market rally also helped the NASDAQ but its bearish topping pattern remains intact for now

The market rally also helped the NASDAQ but its bearish topping pattern remains intact for now


I have adjusted the trading call slightly in the wake of this action. I am maintaining a hold for both bears and bulls, but noting that both sides of the fence need to be prepared for wide swings the rest of this week. I am not going to try to “out think” this one and just hope that I can react quickly and appropriately as the action reveals itself.

I forgot to mention in the last T2108 Update that I decided to go back to buying puts on ProShares Ultra VIX Short-Term Futures (UVXY) because of my perception (I still have not confirmed with the data!) that the Fed tends to dampen market volatility. Of course today’s rally on rate relief smashed volatility and sent UVXY plunging 9.7%. I am still holding a small amount of UVXY shares given September (and October) typically expose market weaknesses. I am definitely dumping the shares by November. In the meantime, trading puts on UVXY remains the big moneymaker.

The iShares MSCI Emerging Markets (EEM) was also a beneficiary of the rally. I was quick on THIS trade and rushed to load up on a fresh tranche of puts as EEM approached 50DMA resistance.


Dreams of a continuation of zero-interest-rate-policy (ZIRP) and a lower dollar even sent international markets higher with emerging leading the way

Dreams of a continuation of zero-interest-rate-policy (ZIRP) and a lower dollar even sent international markets higher with emerging leading the way


As I mentioned, Hilsenrath was even able to move currencies. While the overall dollar index looks like it just swung from one side of a small room to another…


The U.S. dollar in a holding pattern until the Fed bestows its next travelling papers

The U.S. dollar in a holding pattern until the Fed bestows its next travelling papers


…individual currency pairs, especially versus the Australian dollar had notable moves. Here is the 15-minute on the Australian dollar versus the U.S. dollar (FXA). I have stretched out the timeframe to show how, in typical fashion, the currency broke a significant technical level at 0.90 only to eventually shoot higher. There was one more tease for good measure.


The Australian dollar's major technical breakdown does not last long

The Australian dollar’s major technical breakdown does not last long


Given I remain firmly bearish on the Australian dollar, I promptly faded this move.


Through all of the excitement, I cannot forget what happened to Apple (APPL).

AAPL gapped down with the market, but its loss at the lows was much more significant, around 1.5%. Like the S&P 500, AAPL shot up immediately off the low. Unlike the general market, it could not manage to close with a gain. AAPL’s ominous behavior just got more ominous.

AAPL has now twice retested the all-time highs set before the September 9th product announcement and failed. This gap down essentially confirms some kind of topping action. As I have said in previous posts, I remain a fan of AAPL, but I think this poor trading action needs to be taken seriously. Only a fresh all-time high can invalidate the bearish reading. The Apple Trading Model (ATM) predicted an up day, but I was much too slow to get in at the great prices on call options at the lows (recall that the best trades on the ATM occur when the stock opens contrary to the prediction). Eventually I decided to match up calls and puts because it seems that AAPL is going to make at least one more big move this week in one direction or another.


It still looks like some kind of top in Apple (AAPL)

It still looks like some kind of top in Apple (AAPL)


Daily T2108 vs the S&P 500

Black line: T2108 (measured on the right); Green line: S&P 500 (for comparative purposes)
Red line: T2108 Overbought (70%); Blue line: T2108 Oversold (20%)


Weekly T2108
Weekly T2108
*All charts created using
freestockcharts.com unless otherwise stated
Related links:
The T2108 Resource Page
Expanded daily chart of T2108 versus the S&P 500
Expanded weekly chart of T2108

Be careful out there!

Full disclosure: net short Australian dollar, long AAPL calls and puts, long EEM puts, long UVXY shares and puts

Sep
15

T2108 Update (September 15, 2014) – Another Follow-Through on Divergence…But This Time Is Different

written by Dr. Duru
Bookmark and Share

(T2108 measures the percentage of stocks trading above their respective 40-day moving averages [DMAs]. It helps to identify extremes in market sentiment that are likely to reverse. To learn more about it, see my T2108 Resource Page. You can follow real-time T2108 commentary on twitter using the #T2108 hashtag. T2108-related trades and other trades are posted on twitter using the #120trade hashtag)

T2108 Status: 34.7%
VIX Status: 14.1%
General (Short-term) Trading Call: Hold (Aggressive traders can short with tight stops)
Active T2108 periods: Day #309 over 20% (includes day #280 at 20.01%), Day #25 over 30%, Day #2 under 40% (underperiod), Day #5 under 50%, Day #8 under 60%, Day #49 under 70%

Reference Charts (click for view of last 6 months from Stockcharts.com):
S&P 500 or SPY
SDS (ProShares UltraShort S&P500)
U.S. Dollar Index (volatility index)
EEM (iShares MSCI Emerging Markets)
VIX (volatility index)
VXX (iPath S&P 500 VIX Short-Term Futures ETN)
EWG (iShares MSCI Germany Index Fund)
CAT (Caterpillar).

Commentary
Another quiet divergence, another follow-through. Except THIS time is different.

If you only looked at the S&P 500 (SPY), you would think Monday was a surprisingly boring day. In fact, you might think that it was a bullish day given buyers picked up the index off 50-day moving average (DMA) support.


The S&P 500 bounces near 50DMA support to close essentially flat on the day - a nice hammer pattern

The S&P 500 bounces near 50DMA support to close essentially flat on the day – a nice hammer pattern


But next you might wonder why the volatility index, the VIX, crept higher by 6% to one-month highs.


The VIX gains a little again and hits a 1-month high

The VIX gains a little again and hits a 1-month high


So next you look at the NASDAQ (QQQ) and realize that the potentially bullish bounce for the S&P 500 is likely wiped out completely by the very bearish confirmation of a top for the tech-laden index.


The NASDAQ confirms the bearish engulfing top - setting up a critical retest of 50DMA support

The NASDAQ confirms the bearish engulfing top – setting up a critical retest of 50DMA support


At this point, your instincts should point you to trusty ol’ T2108, the percentage of stocks trading above their 40DMAs…if you have not already done so. Lo and behold, you find T2108 dropped another sizable amount, this time by 9.5% to close at 34.7%. This means that under the surface of the market, a bunch of stocks are still sliding rather dramatically. Indeed, as I continue with the chart review, you will see a sample of some of the carnage from today’s selling. Momentum and otherwise highly-valued stocks got hit particularly hard today.

This time is different from the other two examples we have seen of important T2108 vs S&P 500 divergences is that the immediate resolution was only in select stocks. If you had focused on shorting the S&P 500, you would have missed the best plays.

Looking for some positive signs, regular readers know at this point that a large 2-day decline in T2108 generates a “quasi-oversold” trading condition. The 2-day decline of 29.2% has been swift and dramatic. Unfortunately, instead of predicting a bounce, the T210 Trading Model (TTM) is predicting high odds (71%) of more selling (on the S&P 500) for Tuesday. The only solace for the bullish minded is that the classification error is extremely high for the model (45%). However, if you look just at the historical fit, the error rate for a prediction of a loss (or a gain) is only 22%. Consider it a toss-up filled with caution.

With technology identified as the likely culprit, your chart checks should get into high gear.

I checked some of the usual suspects for confirmation that the market attacked with ferocity: Renaissance IPO ETF (IPO), Global X Social Media Index ETF (SOCL), and iShares Russell 2000 (IWM). It just so happens that there are a LOT of potential catalysts this week to motivate selling. For example, there is a Federal Reserve meeting on Wednesday that has the market jittery about hawkishness (oh how quickly the narratives change!), maybe Chair Janet Yellen even takes aim at the valuations of small-caps and social media stocks. A blockbuster IPO from Alibaba (BABA) likely has big traders and investors cashing out of big name, highly liquid stocks to pay for the shares of the century. And there is a triple-witching (or is it quadruple?) options expiration on Friday. In forex, the sudden weakness in the (formerly) stubborn Australian dollar is confirming bearishness all around.


Small caps are right back where the Fed left them...

Small caps are right back where the Fed left them…

A double-top for Renaissance IPO ETF (IPO)?

A double-top for Renaissance IPO ETF (IPO)?

Global X Social Media Index ETF (SOCL) plunges 3.9% for a critical test of converged 50 and 200DMAs

Global X Social Media Index ETF (SOCL) plunges 3.9% for a critical test of converged 50 and 200DMAs


The list of stocks that I saw with massive losses are too much to chart. I wish I could. I just enough time for three down stocks of note, and one upbeat one.

Baidu (BIDU)
If internet stocks sold off because of a desire to move cash into BABA, then BIDU seems like a great choice to play a snapback. BIDU lost 3.3% while Google lost a small fraction of a percent. BIDU cracked its 50DMA and barely closed above support. I speculated ahead of waiting for confirmation with a fresh trade on the stock. The biggest caution is that BIDU made a fresh post-earnings low. One last tranche of buying makes sense if the stock closes the post-earnings gap around $206.


Baidu (BIDU) plunges and barely holds 50DMA support

Baidu (BIDU) plunges and barely holds 50DMA support


Twitter (TWTR)
Twitter is bouncing around the critical $50 level again. It is still above the newly formed 200DMA, so technically it is bullish. However, I think the struggles at $50 are bearishly ominous. TWTR is ripe for a fresh sentiment analysis given the impressive recovery from the point where I estimated negative sentiment had become overdone. I also later noted analysts stepping in to finally turn bullish.

Sorry about the mess of words in the chart below, but I want to keep my notes for future reference!


Twitter (TWTR) gyrates around the critical $50 point

Twitter (TWTR) gyrates around the critical $50 point


Solar City (SCTY)
This triple-top pattern is enough said.


Ominous triple-top kind of action for Solar City (SCTY)

Ominous triple-top kind of action for Solar City (SCTY)


Best Buy (BBY)
Finally, on the positive side, BBY managed to push higher. I am rarely so patient with a stock play anymore. Back in mid-January, I insisted that buying BBY made sense as a contrarian play on a sudden reversal in sentiment that itself seemed so bullishly optimistic in a contrarian way for most of 2013. (A double contrary makes a buy?). I sold a March put that worked out. I also bought what was at the time a long-term (or LEAP) call spread, Jan 27/35 with the assumption that $35 seemed like a good price target. Eight months later, the stock finally lunged at the $35 target, and I sold my position.


Best Buy (BBY) breaks out as it ever so slowly reverses January's massive gap down

Best Buy (BBY) breaks out as it ever so slowly reverses January’s massive gap down


This is a week to stay on your toes and refresh your price alerts!

Daily T2108 vs the S&P 500

Black line: T2108 (measured on the right); Green line: S&P 500 (for comparative purposes)
Red line: T2108 Overbought (70%); Blue line: T2108 Oversold (20%)


Weekly T2108
Weekly T2108
*All charts created using
freestockcharts.com unless otherwise stated
Related links:
The T2108 Resource Page
Expanded daily chart of T2108 versus the S&P 500
Expanded weekly chart of T2108

Be careful out there!

Full disclosure: net short Australian dollar, long BIDU call spread, long TWTR calls and puts,

Sep
12

T2108 Update (September 12, 2014) – Another Quiet Divergence

written by Dr. Duru
Bookmark and Share

(T2108 measures the percentage of stocks trading above their respective 40-day moving averages [DMAs]. It helps to identify extremes in market sentiment that are likely to reverse. To learn more about it, see my T2108 Resource Page. You can follow real-time T2108 commentary on twitter using the #T2108 hashtag. T2108-related trades and other trades are posted on twitter using the #120trade hashtag)

T2108 Status: 38.3%
VIX Status: 13.3%
General (Short-term) Trading Call: Hold (Aggressive traders can short with tight stops)
Active T2108 periods: Day #308 over 20% (includes day #280 at 20.01%), Day #24 over 30%, Day #1 under 40% (underperiod), Day #4 under 50%, Day #7 under 60%, Day #48 under 70%

Reference Charts (click for view of last 6 months from Stockcharts.com):
S&P 500 or SPY
SDS (ProShares UltraShort S&P500)
U.S. Dollar Index (volatility index)
EEM (iShares MSCI Emerging Markets)
VIX (volatility index)
VXX (iPath S&P 500 VIX Short-Term Futures ETN)
EWG (iShares MSCI Germany Index Fund)
CAT (Caterpillar).

Commentary
On July 30th, I pointed out a “quiet” divergence between a plunging T2108 and a relatively calm S&P 500 (SPY). The next day, the S&P 500 took the plunge to the tune of a 2% loss. On Friday, September 12 another quiet divergence occurred.

Just like the last quiet diverge, T2108 has been steadily dropping while the S&P 500 has managed to trade relatively flat. This time around the divergence is particularly wide…and thus even more concerning. T2108 dropped from 49% on Thursday to 38%. Another big drop on Monday, and the market will hit quasi-oversold conditions! Under 40% with this kind of speed also has me wondering about an imminent visit toward true oversold conditions (below 20%).

T2108 plunges

T2108 plunges

While T2108 demonstrates that a good number of stocks are losing altitude again, the S&P 500 has barely responded. On Friday, it only lost 0.60%.

The S&P 500 has lost its primary upward momentum, but it has done well to hover above its 50DMA

The S&P 500 has lost its primary upward momentum, but it has done well to hover above its 50DMA

So if history repeats, we should brace for a notable pullback on Monday. Note well that the Federal Reserve announces another policy decision on Wednesday the 17th. I would normally say this meeting should provide more Novocaine for any of the market’s pains. However, there is some nagging fear that the Fed is going to come out the next meeting with guns blazing with hawkishness. It is amazing how quickly the market narrative can change. Just three months ago the (media-hyped up) fear was that the Fed was falling behind the inflation curve. I was quite skeptical at the time…and rightfully so.

While we wait for the resolution of the quiet divergence and then the Fed meeting, here are some interesting stock charts to study…

iShares MSCI Emerging Markets (EEM)
EEM was my great trade of the week. I mentioned the setup in the last T2108 Update. I almost got a double on the put options. I sold because I did not see any follow-through after the first 30 minutes of selling. With expiration coming up next Friday (September 19th), I thought it made sense to lock in the profits now before time, and perhaps the Fed, erode the value of the puts. These profits easily paid for the calls I bought for a hedge as well as leave me with a tidy net profit. So, if EEM retains its uptrend, I have some time to make money on both sides of the trade while letting the house pay for the bullish side. However, I think EEM’s slide – led by its Brazilian component – confirmed the flag is waving a bright red color.

Despite the small breakdown, EEM is still in an uptrend defined by the 50DMA

Despite the small breakdown, EEM is still in an uptrend defined by the 50DMA

Schlumberger Limited (SLB)
The oil complex is falling fast. Oil-related plays were the darling of 2014. They are quickly turning into goats. SLB has gone from about a 28% gain for the first half of 2014 to a complete reversal of the large surge on June 25th, to a failed retest of its 50DMA as resistance, to what now seems a near certain retest of 200DMA as support. I have this one on my active radar now!

Schlumberger Limited (SLB) looks set to retest its 200DMA support

Schlumberger Limited (SLB) looks set to retest its 200DMA support

Crocs (CROX)
CROX has quickly gone from bullish promise to bearish disappointment. The essential confirmation came last Thursday when a single analyst downgrade to neutral seemed sufficient to send the stock gapping down below 200DMA support. This move confirms a complete wipe-out of July’s large earnings surge (which I used to unload my last CROX) position) and the likely continuation of the downtrend in place since the large gap up on December 30th. Given THAT move failed to erase an earnings gap down from July, 2013, CROX should likely be treated like a broken stock.

The analyst-driven gap down confirms a bearish reversal of the 1-day post-earnings excitement in July

The analyst-driven gap down confirms a bearish reversal of the 1-day post-earnings excitement in July

Logitech (LOGI)
LOGI is having its own problems maintaining altitude through earnings announcements. The chart below does not show the year-to-date flat performance for LOGI that follows anincredible ramp up from 2013: from $7 to $17 from July, 2013 to March, 2014. The chart says it all in terms of fading momentum relative to earnings.

Logitech (LOGI) completes a reversal of the last earnings-inspired gap up - likely confirming waning momentum

Logitech (LOGI) completes a reversal of the last earnings-inspired gap up – likely confirming waning momentum

Monster Beverage Corporation (MNST)
Finally, MNST is facing a tremendous Bollinger Band (BB) squeeze. This one is already on my radar since I am already trying to play a return to upward momentum post-deal with Coke. I am still waiting…

A monster Bollinger Band (BB) squeeze for Monster Beverage Corporation (MNST)

A monster Bollinger Band (BB) squeeze for Monster Beverage Corporation (MNST)

Daily T2108 vs the S&P 500

Black line: T2108 (measured on the right); Green line: S&P 500 (for comparative purposes)
Red line: T2108 Overbought (70%); Blue line: T2108 Oversold (20%)


Weekly T2108
Weekly T2108
*All charts created using
freestockcharts.com unless otherwise stated
Related links:
The T2108 Resource Page
Expanded daily chart of T2108 versus the S&P 500
Expanded weekly chart of T2108

Be careful out there!

Full disclosure: long put and call options on EEM, net short Australian dollar, long USD/TRY

Sep
12

Accumulate McDonald’s

written by Dr. Duru
Bookmark and Share

On September 9, 2014, McDonald’s (MCD) produced yet another awful monthly sales report:

“McDonald’s Corporation today announced that global comparable sales decreased 3.7% in August. Performance by segment was as follows: U.S. down 2.8%; Europe down 0.7%; Asia/Pacific, Middle East and Africa (APMEA) down 14.5%.”

In the U.S. the problem is sluggish industry growth paired with intense competition. In Europe it is poor performance in Russia (surprise, surprise) and weak consumer sentiment. In Asia, MCD is dealing with major supplier issues in China. The Wall Street Journal blames millennials, but I am skeptical of this explanation. The bad news completely overshadowed an announcement later in the day that McDonald’s will accept Apple Pay throughout the U.S. Or perhaps this news helped propel MCD to an apparent bottoming the following day?

MCD closed on September 9th at a fresh 21-month low with a loss on the day of 1.5%. The next day, MCD gapped marginally lower, traded marginally downward from there, and then the buyers took over. On volume of 12.0M shares – more than double the daily average for the past three months and much higher than the 9.7M shares traded on September 9th , MCD closed the day with a GAIN of 2.1%. This 2-day whiplash created what is known as a bullish engulfing pattern: a common bottoming pattern as buyers finally swoop in to pick up shares on the “cheap”, the weakest hands are finally washed out, sellers are exhausted, and bulls and “value” players begin to reverse negative sentiment. In other words, this is a contrarian trade here…


McDonald's (MCD) finally finds some surer footing as buyers rush in

McDonald’s (MCD) finally finds some surer footing as buyers rush in


It is not easy to find such an emphatic bottoming pattern. So when I find one, I get excited. I am as bullish as I can get on this one by beginning a phase of aggressive accumulation of a position in MCD. I am starting with long-term call options since I do not expect a quick and easy road back to all-time highs. After all, MCD is NOT a momentum stock despite a decades long rise (see chart below). With a 3.4% yield, I am very interested in buying shares as well but will wait out the bottoming pattern a little longer: stiff resistance is looming overhead with a declining 50-day moving average (DMA). Traders have a very clear stop sitting right below the intraday low for the year. If sellers manage to push MCD that low again, the bullish engulfing pattern is invalidated. I rate those odds to be very low. MCD seems like a great risk/reward buy here.


A near relentless uptrend over the decades

A near relentless uptrend over the decades


Source for charts: FreeStockCharts.com

Sep
9

T2108 Update (September 9, 2014) – Red Flags Get Brighter

written by Dr. Duru
Bookmark and Share

(T2108 measures the percentage of stocks trading above their respective 40-day moving averages [DMAs]. It helps to identify extremes in market sentiment that are likely to reverse. To learn more about it, see my T2108 Resource Page. You can follow real-time T2108 commentary on twitter using the #T2108 hashtag. T2108-related trades and other trades are posted on twitter using the #120trade hashtag)

T2108 Status: 50.0%
VIX Status: 13.5%
General (Short-term) Trading Call: Hold (Aggressive traders can short with tight stops)
Active T2108 periods: Day #305 over 20% (includes day #280 at 20.01%), Day #19 over 40% (overperiod), Day #1 under 50% (underperiod – minus rounding), Day #4 under 60%, Day #45 under 70%

Reference Charts (click for view of last 6 months from Stockcharts.com):
S&P 500 or SPY
SDS (ProShares UltraShort S&P500)
U.S. Dollar Index (volatility index)
EEM (iShares MSCI Emerging Markets)
VIX (volatility index)
VXX (iPath S&P 500 VIX Short-Term Futures ETN)
EWG (iShares MSCI Germany Index Fund)
CAT (Caterpillar).

Commentary
First there was an ominous print that flagged the end of upward momentum. Next was a close call that demonstrated stubborn resilience again for the buyers. Today, September 9, 2014, the market has swung back into a danger zone. The red flags have brightened enough for me to give the (reluctant) go-ahead for aggressive shorts to stake their claims…with tight stops of course.


The S&P 500 takes an ominous downward turn as 20DMA  gets tested for first time since breakout

The S&P 500 takes an ominous downward turn as 20DMA gets tested for first time since breakout

The NASDAQ's primary uptrend officially comes to an end as bearish engulfing top remains in place

The NASDAQ’s primary uptrend officially comes to an end as bearish engulfing top remains in place

The VIX did not retest recent lows and is now positioned for a new run-up

The VIX did not retest recent lows and is now positioned for a new run-up


The confirming part of these bearish signals is T2108. It closed a fraction under 50% for the first time in three weeks. Its large drop, 11.4%, confirms the end of its run-up and the divergence that I did not dare call bearish last week. T2108 is now leading the S&P 500 downward.

The currency market is providing some additional confirmation for me. Just as I pointed out that the last month of the U.S. dollar’s rally made little to no progress against high-yielding currencies, the U.S. dollar breaks out against the Turkish lira (USD/TRY), and, even more importantly, the Australian dollar breaks down against the U.S. dollar. Moreover, the Australian dollar’s run-up against the Japanese yen (AUD/JPY) has likely come to a definitive end.


The U.S. dollar breaks out against the Turkish lira to a new 6-month high

The U.S. dollar breaks out against the Turkish lira to a new 6-month high

The Australian dollar FINALLY breaks down and confirms the topping action

The Australian dollar FINALLY breaks down and confirms the topping action

The Australian dollar even manages a sharp reversal against the Japanese yen which itself has considerably weakened against all major currencies over the past month or so

The Australian dollar even manages a sharp reversal against the Japanese yen which itself has considerably weakened against all major currencies over the past month or so


Regular readers know I have looked to the Australian dollar versus the Japanese yen for confirmation of bullish and bearish changes in the market. I see here a confirmation of a bearish change. While I am not yet comfortable taking a short position against the S&P 500 (SPY) here, I did rush to buy a hedged options position in iShares MSCI Emerging Markets (EEM) with a heavy bias on put options. EEM has been surprisingly calm during the U.S. dollar’s run-up. It is still in an uptrend as defined by the 50DMA. However, the currency signals above and the increasing volatility in the currency markets in general tell me to expect greater odds of a significant pullback in EEM. The current pricing for EEM options suggests that the market thinks the relative calm will continue, so I like this trade even more since I can buy a potential pullback on the cheap.


Can the relative calm persist for EEM?

Can the relative calm persist for EEM?


To reiterate, the sentiment and momentum is slowly but surely swinging away from the bulls and buyers. The market is not outright bearish, but all the signs for caution are there. Aggressive traders can certainly feel a lot more comfortable going into short positions with well-defined stops. Look out in particular if the S&P 500 breaks through its 50DMA support again. Let’s also not forget that September is one of the most “dangerous” months of the year.

I must end this stock market update with another chart review of Apple (AAPL). The company had its big product announcement today. Despite trading history that tells us NOT to buy into a product announcement, traders did it anyway. The end result was a very wild trading day with large swings in the stock that ultimately settled out into an effective stalemate above 50DMA support. I still think the big sell-off last week ahead of this announcement was ominous enough to take seriously the potential for a more sustained pullback in Apple’s stock. Only a trade to new all-time highs can erase that bearish signal.


A wild ride from one announcement milestone to the next

A wild ride from one announcement milestone to the next

A stalemate, for now, just above 50DMA support

A stalemate, for now, just above 50DMA support


Of course, AAPL’s gyrations just as the NASDAQ (QQQ) is hitting a topping pattern makes the red flags shine all the brighter…

Daily T2108 vs the S&P 500

Black line: T2108 (measured on the right); Green line: S&P 500 (for comparative purposes)
Red line: T2108 Overbought (70%); Blue line: T2108 Oversold (20%)


Weekly T2108
Weekly T2108
*All charts created using
freestockcharts.com unless otherwise stated

Related links:
The T2108 Resource Page
Expanded daily chart of T2108 versus the S&P 500
Expanded weekly chart of T2108

Be careful out there!

Full disclosure: long put and call options on EEM, net short Australian dollar, long USD/TRY

Sep
8

Going Beyond the January Barometer: A More Nuanced View of Monthly Versus Yearly S&P 500 Performance

written by Dr. Duru
Bookmark and Share

(This is an excerpt from an article I originally published on Seeking Alpha on March 3, 2014. Click here to read the entire piece.)

With February providing the lift the S&P 500 (SPY) needed to reverse January’s losses, it is time to take a look once again at the January barometer. The big fear last month was that this barometer indicated 2014 was headed for losses. I demonstrated that the January barometer is actually not useful for predicting down years. {snip}

{snip}

The bubble chart below enhances the chart I showed in the last piece on the January barometer. Here, I have layered in the performance of the subsequent February using bubble size to indicate the magnitude of the change in the S&P 500 for a given February and using color to indicate the direction of the change – white for negative (an unfortunate limitation of Microsoft Excel), green for positive. {snip}


In very select scenarios, February can add some clarity to the likely direction of the year

In very select scenarios, February can add some clarity to the likely direction of the year


As bad as January is at predicting down years, it turns out the remaining months fare worse, sometimes a lot worse. {snip}


January is the best month for predicting a negative year for the S&P 500

January is the best month for predicting a negative year for the S&P 500


While no pattern exists amongst the months to signal the high likelihood of a down year, down years are distinct from up years in the number of negative monthly performances. {snip}



Source for price data in all charts: Yahoo Finance

1974 sticks out as the most brutal – all but one month performed in the red. Otherwise, the pattern seems to be a streak of negative performances occurs in the beginning of the year or in the middle. {snip}

Overall, I think these various nuanced views add more color than the simplistic January barometer. There is of course still no magic crystal ball, but there are interesting monthly patterns to reference throughout the year.

Be careful out there!

(This is an excerpt from an article I originally published on Seeking Alpha on March 3, 2014. Click here to read the entire piece.)

Full disclosure: long SSO puts

Sep
8

Hints of Bottoming for Twitter

written by Dr. Duru
Bookmark and Share

(This is an excerpt from an article I originally published on Seeking Alpha on May 14, 2014. Click here to read the entire piece.)

I last wrote about Twitter (TWTR) three months ago to conclude that the “crescendo top” in the stock was confirmed and to review the importance of sentiment in determining the fate of the stock. I now see a high probability that a crescendo bottom has occurred for TWTR, placing another bookend on the trading opportunity in the stock. In this piece, I update the sentiment picture on TWTR and provide its potential implications.

{snip} The reason’s for some optimism include: a powerful combination of strong revenue growth, especially internationally, and a decline in variable costs; and potential ad revenue growth helped by healthy timeline metrics. I add to that list continued product innovation.

Let’s now see how sentiment has changed for Twitter…

Analyst sentiment
{snip}

Right after TWTR’s last post-earnings sell-off, Deutsche Bank and Goldman Sachs bravely reiterated their buy ratings. Deutsche also lowered its price target from $65 to $52. Since these were rating reiterations, they provide little new information for sentiment.

More interesting are the upgrades that occurred after the disastrous end of the last lock-up period. {snip} Until some analysts start downgrading the stock, I consider the analyst-driven pivot underway for TWTR.

Open interest put/call ratio
{snip}


Twitter's open interest put/call ratio is on the decline again

Twitter’s open interest put/call ratio is on the decline again


Source: Schaeffer’s Investment Research

The $50 bet and momentum
{snip}

The $50 point was not just important as a round number, but it also marked the intraday high on TWTR’s first day of trading. {snip}

What I could not foresee three months ago was that the stock market overall would remain bullish while the market for stocks like TWTR would collapse. {snip}

Lock-up expiration
TWTR’s lock-up expiration on May 6th unleashed a 2-day flurry of selling. {snip} However, just as an extreme surge in buying helped mark a top for TWTR back in December as the market ran out of its most eager buyers, I believe an extreme in selling has helped make a case for a bottom in TWTR as the market may have run out of the majority of its most eager sellers.

{snip}


Twitter attempts to bottom

Twitter attempts to bottom


Source: FreeStockCharts.com

Short interest
Short interest may be the last piece of the puzzle. Bears have yet to release the pressure. {snip}


Shorts have steadily increased the pressure on Twitter. Sentiment now hinges on their next move.

Shorts have steadily increased the pressure on Twitter. Sentiment now hinges on their next move.


Source: Schaeffer’s Investment Research

Given the current data, I am trading for a bottom in TWTR…{snip}


Twitter sentiment on the upswing

Twitter sentiment on the upswing


Source: StockTwits

Be careful out there!

(This is an excerpt from an article I originally published on Seeking Alpha on May 14, 2014. Click here to read the entire piece.)

Full disclosure: short TWTR put spread and puts, short covered call

Sep
8

The Federal Reserve Uses Small-Cap, Social Media Stocks to Display Vigilance on Valuations

written by Dr. Duru
Bookmark and Share

(This is an excerpt from an article I originally published on Seeking Alpha on July 16, 2014. Click here to read the entire piece.)

On July 15th, the Federal Reserve released its latest Monetary Policy Report as a part of Chair Janet Yellen’s testimony before the Senate’s Committee on Banking, Housing, and Urban Affairs. The report covers recent econonmic and financial developments, monetary policy, and a summary of economic projections. The report includes a sidebar on financial stability in the U.S. The Federal Reserve used this section as an opportunity to identify parts of the stock market showing stretched valuations:

“Equity valuations of smaller firms as well as social media and biotechnology firms appear to be stretched, with ratios of prices to forward earnings remaining high relative to historical norms.”

Leading into this sidebar, the Fed appeared to take a swipe directly at social media and biotechnology stocks (emphasis mine):

{snip}

{snip}


Small cap stocks tumble to a new closing low for the month as the Fed takes aim

Small cap stocks tumble to a new closing low for the month as the Fed takes aim

Social media stocks cringe in response to Fed's swipe at them

Social media stocks cringe in response to Fed’s swipe at them


Valuations in many social media stocks are indeed high. A month ago, I looked at valuations of internet-related companies based on price-to-sales ratios, some of which fall into the small-cap or “smaller firm” category for social media, to point out that the Priceline Group (PCLN) deal for Open Table (OPEN) helped to calibrate valuations for this space. {snip}


YELP reverses most of its gains from post-deal excitement

YELP reverses most of its gains from post-deal excitement


Source for all charts: FreeStockChart.com

I also mention YELP because CNBC’s Jim Cramer used the stock to highlight the Fed’s statement on social media valuations in “Cramer: Janet Yellen saying short Yelp?“. Cramer recommended that the Fed target margin rates to extremely high levels rather than target specific stocks to bring valuations down. However, I think he hit the wrong point. The Federal Reserve was not making a call to action. {snip}

(This is an excerpt from an article I originally published on Seeking Alpha on July 16, 2014. Click here to read the entire piece.)

Be careful out there!

Full disclosure: no positions

Sep
8

Campbell Soup: How to Identify A Post-Earnings Recovery

written by Dr. Duru
Bookmark and Share

(This is an excerpt from an article I originally published on Seeking Alpha on March 2, 2014. Click here to read the entire piece.)

After Campbell Soup (CPB) experienced a nasty tumble following its November earnings report, I took a look at the stock as a potential buy for a post-earnings recovery. After reviewing the numbers, I concluded the stock was too expensive to warrant the risk. With some regret for dropping CPB from my radar, sprinkled with a generous helping of hindsight, I have learned some lessons about missing a post-earnings recovery. As it turned out, CPB made a perfect bottom after its post-earnings collapse and went on to print two buyable rallies. After reporting earnings on February 14th that reaffirmed guidance, CPB has printed a healthy 12% gain in just three months. This compares favorably to the S&P 500’s 3.2% gain over the same timeframe.

{snip}


Campbell Soup recovers smartly from November's post-earnings collapse

Campbell Soup recovers smartly from November’s post-earnings collapse


Source: FreeStockChart.com

Downgrade by major analyst fails to cause additional losses
Ironically, the biggest signal of a likely bottom came within days of my ill-timed piece on CPB. {snip}

Large options activity succeeds in moving the stock
Unusual options activity often turns into a false alarm. When this action DOES manage to move the underlying stock, traders and investors should pay close attention. {snip}

M&A rumors succeed in moving the stock
I never recommend buying into rumors. However, if M&A rumors surface within a larger bullish context (or even better, a good technical setup), I am all ears. {snip}

“Surviving” the earnings confrontation following a disaster
Buying into CPB ahead of its February earnings could have been justified by all of the above signals. On February 3rd, CPB took a sharp tumble yet sellers failed to follow-through. This was one last signal for buying CPB at cheaper prices. {snip}

Since the CPB miss, I have written on other post-earnings recoveries that have worked out. So, I have learned some lessons along the way. The main logic I use is as follows: in a bull market, and especially in an economy with positive momentum, companies with solid businesses but a stroke of “bad luck” or otherwise mistimed fortune, present discounted opportunities to play future price appreciation in the general stock market. If the stock market continues higher, a post-earnings recovery should deliver out-performance. If the general stock market takes a tumble, there is (hopefully) a good chance that the recovery stock already has bad news priced into it as a buffer for the pullback.

{snip}

Be careful out there!

(This is an excerpt from an article I originally published on Seeking Alpha on March 2, 2014. Click here to read the entire piece.)

Full disclosure: no positions

Sep
7

The Apple Trading Model (Re)Explained

written by Dr. Duru
Bookmark and Share

The Apple Trading Model (ATM) has become one of my favorite trading tools. I am writing this post to serve as a clear and comprehensive reference point when I write about it in future posts. Interested traders should bookmark this page. (As of writing, my last major AAPL post was another sentiment analysis that ended up anticipating a dip ahead of Apple’s September 9th product announcement).

History
I first thought about an AAPL-related trading model when I realized that AAPL tended to trade higher on Mondays and lower on Fridays. In August, 2012, I wrote “A Guide For Day-To-Day Trading In Apple’s Stock” which laid out my first case showing the surprising regularity in Apple’s day-to-day trading. Almost a year later, I wrote “ATM: The Apple Trading Model Formalized” to apply more systematic machine learning techniques to the analysis. The results produced even more confirmation that AAPL exhibits enough regularities in its trading to make trading worthwhile under certain conditions. The ability to update the model with fresh data represents one of the most powerful features of this approach. The past is a good guide, but it does not guarantee the future.

Since the original formulation, I have made several enhancements to the model as well as refined rules of execution.

The Approach
In machine learning terms, I am using regression (or classification) trees and Bayesian modeling. (For further information on regression trees, see Wikipedia for starters. Ditto for Bayesian classifiers.) The regression trees form the core of the model. I use the e1071 package in R with optimal tuning and 10-fold cross-validation for regression trees. I use the same package for naiveBayes modeling. I only use the Bayesian model as a kind of tie-breaker if the regression trees are producing conflicting and/or inconclusive results, AND I have determined a very good risk/reward trade seems possible. Probably about ninety-percent of the time, I trade just based on the regression tree results. So, I am simplifying the rest of this discussion by focusing on the regression trees.

The ATM model classifies based on whether Apple’s stock closes on a particular day up (1) or down (0). I do not derive any relationship to explain the magnitude of AAPL’s change on the day. Such a model would of course be a lot more powerful, but I have yet to derive one that produces acceptable error rates. I compensate for the binary classification by using options: options can make a good amount of money on small changes and losses are automatically capped. Options are also appropriate since the trading decisions are short-term (daily).

I keep the model simple by using a small set of factors: the day of the week, the month of the year, the previous day’s change in value for the S&P 500, the previous day’s change in price for AAPL measured from that day’s open, the previous day’s change in price for AAPL measured from the previous close, and the number of days AAPL has closed up over the last 5 days (a short-term trending factor). Note that the assumption of independence across all these factors is a strong one, making the Bayesian model inconsistent (aka perplexing!) in its behavior at times.

For data, I use daily data starting from 2010. This choice is based on the assumption that the most recent data matters the most for short-term trading. The years of 2008 and 2009 were so unique to the history of trading, I prefer to exclude them from models that look for regular patterns. I use daily data for both Apple and the S&P 500. The S&P 500 provides the general market context. I do not use the NASDAQ because it is more representative of a single sector of the market (high-tech). You can download the latest historical data by clicking this link.

Model setup
Here are the independent variables (or factors) used in the Apple Trading Model. You must understand these to read the model output:

  • DayOfWeek: Day of the week – 2 to 6 is Monday to Friday.
  • MonthOfYear: Month of the year – 1 to 12 is January to December.
  • SP500PrevDailyChg: Previous day’s performance for the S&P 500 (SPY) (daily percentage change). “Previous” means the trading day prior to the day being classified (projected).
  • AAPLPrevIntraDayChg: Previous day’s performance for AAPL (daily percentage change) measured from the OPEN of the day.
  • AAPLPrevDailyChg: Previous day’s performance for AAPL (daily percentage change) measured from the CLOSE of the prior day.
  • AAPLPrev5Days: Number of days AAPL closed up out of the last five trading days – a measurement of short-term trending.

Execution
I run the ATM for a set of years starting with the first year of historical data (2010). As of 2014, this means four passes: history based on data from 2010 to 2014, 2011 to 2014, 2012 to 2014, 2013 to 2014, and just 2014. I run these sets twice for a total of 8 runs: four for the classification for AAPL’s expected close measured from the previous day’s close – I call these the “Closing Sub-Models”; four for the classification for AAPL’s expected close measured from the trading day’s open – I call these the “Opening Sub-Models”. The trading day is the future day targeted for a trade.

This breakdown allows me to detect changes in AAPL’s trading behavior over time. It also allows me an opportunity to check whether I should trust the conclusions based on the latest year of trading. I compare and contrast each sub-model’s classification error as well as the historical frequency for the given trading scenario. I trust most the sub-model using the current and previous year’s data. Of course, I prefer to see the sub-models agree as much as possible with each other. I dial up or dial down my trading aggressiveness based on the degree of agreement across the sub-models.

So far, so good on classification errors: I am OK with anything below 35% or so. As of Friday, September 5, 2014, here are the classification errors listed by starting year of the sub-model. The first number is the classification error for the Closing Sub-Model, and the second number is the classification error for the Opening Sub-Model.

2010
0.2486393
0.2273215
2011
0.2482028
0.2217553
2012
0.2663185
0.2375038
2013
0.2745552
0.2471611
2014
0.3192823
0.2650851

In my opinion, the classification errors are remarkably good across the board. Granted, I am surprised that the 2014 Closing Sub-Model has the highest classification error now. These results mean all the sub-models are quite useful and functional – a very good, even if surprising, outcome!

Here are links to the charts of the regression trees listed by the first year of historical data used in the model. These trees define the decision rules for making a trade for a given day. The terminal nodes provide the historical frequency of the given scenario described by the path to that node. Following a condition to the left represents “TRUE.” Following a condition to the right represents “FALSE.” Again, use the key provided above to understand the labeling. I update these charts on a periodic basis.

Trade from the previous close – Closing Sub-Models
2010
2011
2012
2013
2014

Trade from the open (intraday) – Opening Sub-Models
2010
2011
2012
2013
2014

As I stated earlier, I strongly prefer to use options for trades using the ATM. I have tinkered with trying to anticipate the outcome of the model and/or run the model based on anticipated closes with mixed results. I now only do such anticipation for Monday and Friday trades. One benefit of frequent model updates is that I am currently reviewing how Friday trades have increasingly become less certain. It is almost as if traders are finally trying to anticipate Monday’s gains by buying into Friday. This of course reduces the potential upside for Monday. Over time, new patterns develop as traders increasingly anticipate old patterns.

I have come to prefer running the model after the close and then looking for an ideal entry the next day. This allows me to opt out of trading if conditions are not right. For example, if the model predicts an up day, and AAPL gaps up or races up to a 1% gain at the open, I will abstain from trading. The risk/reward becomes very poor under such conditions. If on the other hand the model predicts a down day, I might open up a trade to fade, especially if BOTH the Opening and Closing sub-models agree on a down close. In this case, the model may still end up wrong, but I have a chance of catching a fade as AAPL fails to sustain the early gains. Given AAPL has been in a bullish phase since its bottom last year, I have a great preference for buying dips on AAPL in anticipation of upside the following day (in the case of a Friday dip) or by the close.

These are the basic contours of the ATM model. If you have questions or feedback, feel free to post them below. I will definitely get back to your post as soon as possible. Also stay on the look-out for future enhancements. I plan to do a year-end review for overall performance metrics and to analyze what is working and what needs improvement.


A very systematic trading year for Apple

A very systematic trading year for Apple


Source: FreeStockCharts.com

Be careful out there!

Full disclosure: long AAPL call options

 Subscribe Subscribe Subscribe by Email Follow DrDuru on Twitter
Follow DrDuru
20-50% off entire stock NCAA apparel at Kohl's OLD ARCHIVES

Recent Archives

Great Links

Tag Cloud

Join In

Disclaimer