Joined Nov 11, 2007
1,458 Blog Posts

SPY Closed at 25 Day High, Volume Was Low. Bullish or Bearish?

On Monday, SPY closed at a 25 day high, but volume came in less than the 50 day average. The combination of a 25 day high made on low volume seems to be a bearish combination, but one never knows without backtesting it.

For some background, read this post which shows what happens after X day new highs are made.

The Rules:

Buy SPY at the close if

  • SPY closes at a 25 day high AND
  • Volume is <80% of the 50 day average

All SPY history was used and no commission or slippage was included.

The Results:

My suspicions may have been proven to be correct. It appears that SPY does not have enough conviction behind it to hold its gains when a 25 day high is made on lower volume. When the volume requirement is removed (see the red line), SPY has tended to hold onto and continue adding to previous gains.

While 50 days after the setup SPY has managed to show gains, this may be due to the fact that markets have a bias to the upside. What is more important to consider is that the period in between has seen sideways and up and down trading. In other words, while volatility tends to drop as SPY makes new highs, the lack of buyer conviction may mean that volatility is going to remain elevated despite the 25 day high.

Sample size was not an issue with this study.

Interestingly enough, another recent study lead me to conclude that “[t]he results show that we can expect more sideways trading and volatility.

Our current environment contains many uncertainties, but markets have always been able to shrug these things off and continue climbing. However, the studies and the market environment both seem to be in agreement here. Thus, my call for more sideways action and elevated volatility remains.

Comments »

Competing Indicators: One Says Hold Long, One Says Sell

My ROC indicator is still holding on to a long SPY position while Varadi’s AggM indicator has recently flashed a sell.

ROC indicator is the middle pane with the blue and red lines. Varadi’s AggM is the bottom pane with the single red line.

Honestly, the ROC indicator was designed to catch long-term trends. It has seen a lot of whiplash over the past 6 months. For this reason, I’m inclined to put more weight into the AggM.

Comments »

$VIX Trades for 100 Days Below Its 200 Day Average: Bullish or Bearish?

I recently published a simple study which examined how SPY has performed after $VIX closed above its 200 day average. The flip side of that study is to examine how SPY has performed after $VIX has closed for X days beneath its 200 day average and then closes above it.

On June 1st, $VIX closed above its 200 day moving average after trading beneath it for 115 days. There are only six instances in the entire $VIX history where it has traded for more than 100 days beneath its 200 day moving average. Does the number of days it has traded beneath the 200dma have any discernible effect on SPY performance if SPY is bought at the close the day $VIX closes back above its 200dma?

The Rules:

Buy SPY at the close if

  • $VIX has closed beneath its 200dma for more than X days
  • $VIX closes above its 200dma

SPY is sold Y days later. No commissions or slippage included. All SPY and $VIX history used.

The Results:

The results show that we can expect more sideways trading and volatility. While the >99 days results are promising, there were only 5 trades. The dates and results for these 5 trades are below. Perhaps 20 years from now we will have enough samples to draw some conclusions, but for now, we must only observe and try not to make major decisions on the results generated from only 5 trades.

EntryDate      % Gain/Loss

SPY   8/3/1999     -3.21%
SPY    3/10/2004  -2.63%
SPY    2/27/2007   8.36%
SPY    1/22/2010    9.00%
SPY    6/1/2012       3.71%

The >24 days below results were generated from 30 trades. 25 trades were held the full 50 days. While this is not a large sample, I believe it is large enough for us to expect more volatility and sideways trading.

Comments »

I Know You Want to Talk Spanish Bailouts; Let’s Talk Baseball Instead

Look, I published articles showing that it wasn’t a bad idea to get long on the correction, and hopefully you have, what with Spain being bailed out and all. But disregard that. This evening we talk baseball.

And honestly, we are talking baseball because I’ve been at my son’s all-day tournament, where they won 2 and lost 1, losing the championship game 14 – 12. Frankly speaking, I’m worn out and don’t have the mental or physical capacity to do much more than brag on my boy a bit.

He’s been crushing the ball, and had several key hits today. But rather than describing them, I thought I’d just post the video. I have other videos from big hits today, but you can read his name on his jersey and I don’t need internet crazies harassing my family, so one video will have to suffice.

This hit was during the championship game. You can see from the scoreboard they were down 9 -6 at the time.

Yes, that is me with the somewhat too-high-pitched whoop after the hit. You’ll have to trust me that I was hoarse by that point. It is not a true representation of my manly whoop. Don’t miss my littlest making his cameo appearance as bat boy.


Comments »

Be Careful With This Sour $AAPL Strategy

A reader recently sent me a strategy for trading $AAPL and asked if I’d take a look at it. The strategy was published on Seeking Alpha and can be read here: An Apple Story: Alpha, Algos And A Poorman. The strategy is very simple. As I love simple strategies, I coded it up and tested it. Unfortunately, except for a year or so of good performance, running the strategy over any length of time will probably result in complete separation of the trader from his capital.

What is the Strategy?

The author, Mr. Lewis, believes that options are heavily influencing $AAPL trading and he noticed an anomaly which he believes is caused by the high number of $AAPL calls that are traded. If you want more detail, read his explanation.  Anyway, the strategy seeks to exploit the anomaly caused by the options trading.

  1. Buy $AAPL at the close on Friday.
  2. Sell $AAPL at the close on Tuesday.
  3. Short $AAPL at the close on Wednesday.
  4. Buy-to-Cover $AAPL at the close on Friday.

Thus, the strategy is always in the market, except for Wednesdays.

Let’s look at recent (2011 to the present) performance of this strategy:

Not including commissions or slippage, the strategy has returned a compound annual rate of 93.28% with 61.24% of trades winning and a maximum system drawdown of -10.85%. The Sharpe Ratio is 3.28. To put it plainly, those are awesome results. Just buying and holding $AAPL over the same time period resulted in a compound annual rate of 48.02% and a maximum drawdown of -16.58%. The strategy doubles the return and reduces the drawdown by almost a 1/3rd.

What’s not to like? Well, we might want to look and see how the strategy has performed over a longer period of time.

Now let’s look at the results generated from trading the strategy over all of $AAPL history.

Using all $AAPL history the strategy has returned a compound annual rate of -5.16% with 50.19% winning trades and a maximum system drawdown of -94.86%. The Sharpe Ratio is 0.06.

I decided to remove the short requirement from the strategy. The results are below.

Since 2003 the strategy has returned 27.88%. While it looks like the long-only, buy Friday close and sell Tuesday close has kicked butt, buying and holding $AAPL from 2003 to now has resulted in a return of 58.75%. It makes sense that holding a company only 2 of every 5 days while it has been in a massive bull run will result in a lower return.

At the end of his article, Mr. Lewis provides these insturctions:

You cannot decide to trade this some weeks and not others or your results will severely be altered. It is designed to be traded as if you were a robot, there is no thinking. You have set instruction [sic] that should be followed at all times.

Mr. Lewis finishes with his “true goal” for publishing this strategy:

My true goal is for Apple to trade at fair value and unabated. If these artificial waves are going to stay in Apple, I want everyone to make money off of them. They will either become so big that the whole world will start riding them, or they will go away. But, until then, surf’s up dude.

I would be careful with this strategy. Perhaps the author is correct and there is an options anomaly to be exploited. I offer that correlation does not imply causation. My guess is that his strategy has worked the last 1.5 years because the buy Friday, sell Tuesday, short Wednesday, cover Friday approach has been curve fit to recent $AAPL short-term cycles. As the long-term equity curve shows, this cycle has worked before, and then abruptly stopped working. The strategy has also spent almost 9 years doing generating flat returns. If I were going to trade a strategy “as if [I] were a robot,” then I’d want to be sure that returns have been robust over a long period of time. The returns from this $AAPL strategy certainly have not been.


Comments »

Modeling the Bottom: It Bounces and Then Flattens

Apologies for the double entendres. I still have bclund’s recent post on the brain. More seriously, on Monday, June 4th, SPY made a new 102 day low. As you are all well-aware, after making the new low, the market bounced, gaining slightly more than 3% as of today’s close. Tonight’s installment creates a rough model of this pattern, finds previous occurrences of it, and calculates historical performance after it occurs.

The Rules:

Buy SPY at the Close if

  1. it makes a new 100 day low AND
  2. the three-day rate-of-change is greater than 2.99%

SPY will be sold X days later.

No commissions or slippage included. All SPY history used.

The Results:

And below are the trades that were held the full 50 days:

While there were 21 occurrences of this setup, there were only 13 samples held for the full 50 days. Those trades are shown above.

Even though the modeling was a rough approximation and not extremely specific, it still did not yield very many samples. The typical caveats about small sample sizes apply.

I posted the individual trades because I thought it was important to understand the impact of the 2008 trades. If 2008 can be considered an outlier, than the rest of the results are promising.

I’ve heard many traders guessing that the next few days will bring some consolidation and then another pop. Modeling of recent action shows that historically, that has been true.

Comments »

Positive Divergence: Number of Stocks above their 20 and 50 Day Averages

Approximately 2 weeks ago I posted about the bottom signal that was developing, signified by the historic lows of the percentages of stocks above their 20 and 50 day moving averages. Shortly thereafter, a bounce was made, and then more downside action followed.

Click on the chart to enlarge…

I have updated the graph, and as you can see, there is a divergence occurring. Note the percentage of stocks above their 20 and 50 day averages (green and red lines )has climbed while SPY has fallen.

Some of you noted in the previous post that after a historic low was made, the market had the tendency to fall a bit more or consolidate before beginning a new uptrend. While the market continued to fall or consolidate, breadth, as measured by the percentage of stocks above the 20 and 50 day averages, tended to have already bottomed. It appears the same progression is now occurring. Check the previous post  to see a longer-term view of the graph.

I believe the divergence signifies we have better than average chances of a significant bottom developing here.

I should note that I had written the bulk of this post last night, to be published today, and had no way of knowing that we would get a huge bounce today.

Comments »