The Nasdaq 800 Index

The what!?

Yeah, I know, there is no such thing but I'm making one up anyway!

To continue my venture into the world of trading small cap stocks, I have created a new watchlist that I'm calling the Nasdaq 800.

This new small-cap index includes 778 NASDAQ-listed stocks priced between $3 and $25 with average volumes between 100,000 and 1 million shares. This watchlist will be the primary, although not exclusive, source of future trades. I will trade NASDAQ 100 stocks as well.

After market close on Friday, the following two stocks popped up as possible trades.



So far so good. But then this stock was flagged as a buy, too. Ugh, what an Ugly Chart (no offense Mr. Ugly Chart).


I see nothing in this chart to like or love, so I probably won't be hitting the buy button on this one.

It's All Relative

Yesterday I decided to put theory into practice by making three stock trades using the Relative Strength vs. Naz100 screen mentioned in an earlier post.

The system's targets of choice were Simpletech Inc. (STEC), American Oriental Bioengineering (AOB), and Quicksilver Inc. (QZK). Each stock will be held until the system issues a sell signal or until the stoploss is hit.

Position size is determined with the following formula:

Risk Per Trade / (9-day ATR x 3) = Number of Shares

Risk per trade is limited to .5 percent of equity plus commission. Risk will be increased to 1 percent of equity per trade once I'm comfortable with the system.

The trailing stoploss is the product of (9-day ATR x 3) and is put in place as soon as the buy order is filled.

Once the price has moved far enough in my favor, then the trailing stoploss will be replaced with a breakeven stoploss just above my entry point. The goal is to give prices enough room to move without allowing winning trades to become losers.

In many of my tests, the Maximum Favorable Excursion for losing trades shows dozens of trades that were up 10 percent or more before ultimately losing money. If a stock is up 10 percent, then there is no excuse for allowing that stock to wind up a 30 percent loser.

Well, that's my story anyway and I'm sticking to it...for now. If this methodology ends up cutting short too many winning trades then I'll rethink it.

Finally, if a stock's price hits the stoploss without ever turning positive, then I will re-enter the trade if the price closes above the original entry point.

Enough with the rules, here are the charts for Friday's trades:

Reading List

Last night I ordered the newly revised edition of Trade Your Way to Financial Freedom by Van Tharp. I had planned to purchase the book earlier this year, but then learned that an updated version was in the works, so I waited.

Other books I've read this year are listed in this previous post.

So what is your favorite or most useful book(s) on trading and why?

Please post your recommendations in the comments.

Small Caps, Big Profits

On Wednesday, I wrote about a program implementing InvesTools "buy on the green arrows, sell on the red arrows" trading strategy and shared some backtesting results using the program on the Naz 100.

In the comments, Marlyn from Filtering Wall Street suggested I try the InvesTools approach on small cap stocks.

Tonight I took his advice and created a watchlist of 100 stocks that fall on the smaller side of the volume scale. I performed a stock screen at MarketWatch for stocks with daily volume between 100,000 and 1 million shares. I then sorted by volume and placed the 100 highest volume stocks in the watchlist, which produced a list of stocks with daily volumes between 750,000 and 1 million shares.

I then ran backtests using three different programs, including the InvesTools program, my 10, 20, 50 EMA crossover system, and a relative strength/moving average system I have been experimenting with for a couple weeks.

Based upon the results of the three tests, I think I'm going to focus my attention and trading on the smaller fish in the Wall Street ocean!

Promises of Profit

I'm an evening person. In my ideal world, life wouldn't begin until 10 a.m.

When the rest of the family is in bed, I will often stay up until 1 a.m. or later reading a book or researching something on the web. I can stay up all night, just don't ask me to function in the morning. And no, going to bed early doesn't make me more of a morning person. Mornings just suck, they always have.

The TV may be on in those early morning hours, but that's just for background noise, at least until the parade of infomercials starts up with their promises of profit and breathtaking returns in stocks and options. All you have to do, of course, is sign up for their educational programs and follow their proven system. Instant wealth is there for the taking.

But there's no need to sign up for expensive educational programs to find out if the system works, all you need is access to some trading software.

The most popular system being promoted in infomercials these days is from InvesTools, which promises to make technical analysis easy. They promote their system in simple terms: buy when you see three green arrows, sell when you see three red arrows. Simple. Easy. Profitable.

But is it?

Over at Wealth Lab, a user asked if anyone had created a ChartScript based upon the InvesTools methodology, which he described as:

Buy if MACD and Stochastic are positive and the price is above 30-day SMA.

Sell if MACD and Stochastic are negative and the price is below 30-day SMA.

The above may be a bit simplified, but I think it hits the high points of the system's buy and sell triggers.

In less than two hours, another user submitted a ChartScript implementing those buy and sell rules, which only takes a couple minutes to create if you are using the downloadable version of Wealth Lab from Fidelity. Us website users are left pecking out code one line at a time. To read the code, go here.

So, can we make any money using those buy and sell criteria?

The table below shows the results of applying the InvesTools approach to the Naz100 stocks, and compares the results to buying and holding the QQQQ over the same period, which is five years. Position size was set at 10 percent per stock.


Results for the S&P 100 stocks were a less than eye popping but still respectable 80 percent, while applying InvesTool's buy and sell criteria to the Dow 30 generated a measly return of 2.65 percent.

The First 60 Feet

In my younger years, I used to go drag racing on Sunday morning's with a family member and two of our friends. The three of them poured every spare dollar they had into making their race cars go as fast as possible. We were amateurs racing other amateurs and having a grand ol'time doing it.

One of the keys to winning, however, wasn't the car's top speed, but rather how fast you could cover the first 60 feet of that 1/4 mile. The 60-foot split time, quickness at the start, oftentimes determined who won or lost.

In memory of those automotive contests, I present to you the financial equivalent: a head-to-head stock race between Ford and Toyota. Not quite as exciting as roaring engines and the smell of burning rubber and high-octane fuel, but still interesting. Unfortunately, it hasn't been much of race these past three years.


Disclaimer: long Ford.

Why Lose?

Every trading system will generate buy signals that ultimately end with some trades closing out for a loss.

If the trading system has an edge and is consistently executed, and money management rules are adhered to, then losing trades should be mere speed bumps on the road to higher profits and an equity curve graph worthy of framing and hanging above the mantle.

But I hate losing trades.

As I continue to explore the potential and pitfalls of using moving averages to trade individual stocks rather than markets and indexes, I am left wondering why specific stock trades ultimately turn out to be losers.

Is it something about the specific stock that makes a promising moving average crossover turn to a loss? Lower than average volume perhaps? Or is the root cause more broad such as underlying market conditions? And can these factors be identified and reduced to computer code so that an automated trading system can check for these conditions before acting on the buy signal, thus reducing the number of losing trades?

While testing my moving average system, I noticed a couple things. First, moving average crossovers is a profitable way to trade that can outperform buy and hold investing with less capital exposure and smaller drawdowns. Secondly, losing trades tend to clump together rather than being spread randomly throughout the trading calendar.

To begin my quest to root out losers before they are losers, I decided to plot the purchase point of all buy signals issued during testing. I used a weekly chart of the 2006 NASDAQ 100 with moving averages corresponding to the daily moving averages used to generate buy & sell signals, i.e. a weekly EMA of 2 equals a daily EMA of 10, EMA 4 equals EMA 20, and EMA 10 equals EMA 50, assuming five trading days per week. Losing trades are red dots and winning trades are blue dots.


What I found was no surprise. Buying uptrending stocks when the broader market is moving down is a very bad idea. It appears that whenever the weekly EMA 2 is declining or below the weekly EMA 4, then those trades have a tendency to be closed out as losers.

I know, you were expecting something more profound than the tired old cliche "don't trade against the trend".

Sorry to disappoint, but Jesse Livermore was right, don't go long in a bear market.

I will plot trades from other years tonight and then post the resulting charts. In the meantime, I'm going to write a market filter that will eliminate buy signals for uptrending stocks in downtrending markets. I'll share the results when they are available.

Are We There Yet?

No, not to grandma's house, to the market top?

Dr. Steenbarger serves up his latest chart demonstrating The Trajectories of Bull and Bear Markets.

Festooned in holiday-correct green and red squiggly lines, Dr. Steenbarger looks at the historical record of significant buying and selling days to help us determine if the bull is heading for a fall.

As always, a very insightful post worth your time.

Got Leverage?

In my previous post, I shared results from several backtests using different moving averages on the QQQQ and SPY, and while each option outperformed the buy and hold approach, results of all strategies were rather anemic.

A 35-percent return over five years isn't exactly something to brag about, even if it did outperform the 20-percent return of buy and hold while cutting market exposure in half. A three-year declining market from 2000-03 will do that to ya.

So what's the answer?

Why, leverage of course!

(insert shrieks of horror and risk of ruin comments here)

And since I know nothing about trading options or futures (you'll blow out your equity kid!), I went searching for another return boosting tool.

ProShares this summer launched new ETFs that seek to double the return of the NASDAQ 100 and S&P 500 (tickers: QLD & SSO). If the Naz goes up 10 percent, then the ETF should go up 20 percent, at least in theory. This tool allows those of us who are clueless about trading options and futures an opportunity to put them to work for us.

Unfortunately, the ProShares ETFs haven't been around long enough to establish much of a trading history, so when I wanted to test the 10, 20, 50 moving average strategy with a leveraged product, I had to look somewhere else.

What I found was the UltraOTC and UltraBull mutual funds by ProFunds (tickers: UOPIX & ULPIX).

Established in 1997, these funds are designed to double the return (or loss) of the NASDAQ and S&P 500, respectively. With nine years worth of price data on hand, it was time to backtest and here is what I discovered.

Applying the 10, 20, 50 averages to the Ultra funds increased net profit from $17,320.32 to $29,215.29 for a difference of $11,894.97, or a 69 percent increase. Not quite up to the 100 percent target, but still a very substantial increase in returns.

While the return increased by 69 percent, the amount of capital used to obtain those returns decreased from 50.17 percent to 47.42 percent, which helped push the risk adjusted return up to 123.23 percent from 69.05 percent.

On the downside, the maximum drawdown experienced with the Ultra funds increased from a modest 11.85 percent to 23.03 percent, but that was much better than the buy and hold approach, which had a maximum drawdown of 71.32 percent.

After five years, the net profit for the buy and hold approach was -3.74 percent. I guess buying and holding leveraged mutual funds is a bad idea!

But strategically buying and selling seems to be a viable path to higher profits.

10, 20, 50

Spent some time Saturday experimenting with different moving averages over at Wealth Lab. See the table below for the results trading the QQQQ and SPY.

Buying the EMA 20/50 crossover and selling the EMA 10/50 crossunder seems to give the best balance between risk and return, even though buying the 15/50 crossover and selling the 10/50 crossunder provided the highest return by about 1 percent. The former has a better risk adjusted return and a lower drawdown. The differences, though, are not that great so either would work well.

I used a faster moving average for selling because markets go down faster than they go up. Exiting faster seems to increase overall return by more than 5 percent.

All six options outperformed the buy and hold approach.

Read, Research & Test

In preparing for my conversion from fundamental investor to system-based trend follower, I have spent a lot of time reading books on other traders, systems, and techniques. I like to read, so this has been a fun time of learning for me.

So far this year, I have read the following books twice:

• Trend Following by Michael Covel
• Market Wizards and New Market Wizards by Jack Schwager
• Getting Started in Technical Analysis by Jack Schwager
• Reminiscences of a Stock Operator by Edwin LeFevre

I also downloaded and read a copy of the original Turtle Trading rules back when they were available free of charge. I think they can be had now for $20.

Once I had acquired a basic understanding of the concepts behind system-based trend following, I then started poking around the Wealth Lab website. Over at Wealth Lab, you can write and test trading systems using nearly any parameter you can dream up.

One caveat though, writing systems on the website requires an understanding of the PASCAL programming language, but don't let that scare you away. There are more than 1,000 programs that Wealth Lab users have created and posted to the website for your testing pleasure. I'm not a programmer, but I have found it fairly easy to modify existing programs to do what I want. When all else fails, the user forums provide a quick and ready answer to your problem.

Wealth Lab trading programs are called ChartScripts and you test them against Watchlists that you create, usually stocks, ETFs, mutual funds, or indexes. Once your system is ready, use the Scan feature each day to generate buy and sell signals for trading – paper or actual.

One limitation in using the website is its dependency on end of day prices. Performance of your model portfolio is not updated until end of day, and buy or sell signals generated during market hours are based upon the previous day's closing price. On the plus side, you can run your scans during the evening, enter your buy or sell orders with your broker, and then actually work while at work instead of watching the quote screen.

While intraday traders will not be happy with the website, I am finding that the reliance on end of day prices is teaching me patience. I'm trying to be a trend follower, so why do I need to check the account balance of a real or model portfolio during trading hours, or stare at a quote screen all day? The answer, of course, is that I don't need to nor should I. I'm not a daytrader, nor am I trying to become one, so I need to stop acting like one!

Cantor Fitzgerald analyst Derek Brown today issued a buy rating for Google and set a $650 price target.

To be honest, I have no idea how much the company is worth, or where its stock price is headed long term, but a quick look at a chart indicates investors will have to overcome a lot of near-term weakness just to see the other side of $500.

On Nov. 30, the stock closed at $485 bringing its five day EMA below its 15 day EMA. Since diving into technical analysis this year, I have been using the five day EMA to gauge a stock's near-term strength. If the bulls can't keep that average above the 10 or the 15, then support is pretty thin.

From Nov. 30 to Dec. 20, the five drifted below the 15 EMA until yesterday when the five fell below the 50 day EMA. Making matters worse, the 50-day EMA is beginning to turn down.

So will GOOG hit $650? I have no idea, but I do know it is going in the wrong direction right now and that I wouldn't buy until it starts demonstrating something resembling strength.

UPDATE: Marlyn from Filtering Wall Street suggested in a comment that I take a look at a two-year chart using the 21 week and 90 week EMAs. Notice how GOOG bounces off the 21 week EMA then retreats. Click here to see Marlyn's take on GOOG.

Kevin's Market Blog shares some interesting charts demonstrating why he is bearish on this market.

A previous post featured performance data for a dip-buying system that buys when the price of the SPY and QQQQ declined for two conseuctive days, then sold after two consecutive up days. Such a system can be very profitable on a consistent basis. Check out Filtering Wall Street for his take on the system.

Here is a variation on a theme called HiLoLimit, which I also found at Wealth-Lab. The best description for how it works comes from Wealth-Lab member SideShowBob:

Take the highest high from the last x bars, and the lowest low from the last x bars, take the difference between them (high - low) and then place a buy order y % below the band.

Every day a new limit price is generated based on the prior day's price. You then wait until a limit order gets filled, and then set a profit target (~1.7%). There's no stop loss, so the other way to exit a trade is a days expired parameter. This means if the profit target isn't hit you sell at the open on the zth trading day after the trade was entered.

This system issues a lot of limit orders when used with stocks, so automating the placement of buy orders is a good idea. Unfortunately, the system appears to have a fairly low rate of filled trades since you are waiting for the price to go down and touch your entry point, but filled trades are winners 70 percent of the time or more.

An interesting strategy would be to combine this system with a long-only moving average trend following system. Check out the equity curve and notice how HiLoLimit is mostly in cash during long uptrends, when your trending system has you fully invested, but then kicks in during choppy or downtrending markets, when your long-only trending system has you in cash.

I'm not smart enough to take two different systems and combine them into one computer program, but the potential is intriguing.

These performance numbers are generated using the following parameters:

• Trading QQQQ and SPY for last five years
• $50,000 starting capital
• 50 percent of equity per trade
• 1.7 percent profit target
• Exit all trades after 7 days if profit target not hit

Dr. Steenbarger and Yaser Anwar are collaborating to provide investors and traders two sides of a stock's story. Here is the scoop from Investment Ideas by Yaser Anwar:

This is a new feature for this blog & I'm glad to inform you that Dr. Steenbarger aka Mr. TraderFeed & yours truly have agreed to collaborate to provide readers two sides of the same stock. Dr. Brett will provide his usual succinct short-term trade direction and yours truly will dissect a longer term trend.

I thought I would join the fun and share another side, how their stock of choice performs using my long-only trend following system based upon the 15 and 50 EMAs. Buy on the crossover, sell on the crossunder, you can't get more basic.

Here is how AIG performed using $5,000 per trade since December 2004. The results are courtesy of Wealth-Lab.

Dr. Brett Steenbarger at TraderFeed takes a look at mini-trends in the SPY and QQQQ to determine if there are any opportunities to profit from two consecutive up or down days.

His post reminded me of a ChartScript program called Smitlener Streaks that I have tested at the Wealth-Lab website. The program is named after Damir Smitlener who published an article entitled "Where is the Weakest Link" in the December 2004 issue of Stocks and Commodities Magazine.

The concepts in the article were turned into computer code by Robert Sucher of Wealth-Lab. The trading rules are as follows:

1. Buy a small position upon detecting 2 consecutive lower closes.
2. Add a position each time you detect the condition above.
3. Exit all positions upon detecting 2 consecutive higher closes.
4. In case of limited cash, choose the trade with the lowest 14-day RSI

I ran the ChartScript using the SPY only, the QQQQ only, and the SPY and QQQQ combined. For all test runs, starting capital was $50,000. The SPY only and QQQQ only tests used position sizing equal to 50 percent of equity, which allowed the system to double down per rule number 2.

The Combined test used position sizing of 25 percent of capital, which allowed the system to double down in each, or to take three positions in one of the two. Obviously, changing position sizing will dramatically change the results, but here is what I got for these particular parameters.

SPY Only


SPY Equity Curve


QQQQ Only


QQQQ Equity Curve


SPY & QQQQ


SPY & QQQQ Equity

Michael Covel, author of the book Trend Following: How Great Traders Make Millions in Up or Down Markets, spoke at the annual CLSA Investors Forum in Hong Kong back in September. He then posted the following video excerpts from his speech on YouTube, which touch on some of the basic philosophies of trend following.

For more on Covel you can visit his website at www.trendfollowing.com.

In the beginning, there was fundamental analysis with its income statements, balance sheets, discounted cash flows, p/e ratios, book values, and the buy and hold approach to wealth creation.

By far the most popular way to invest in stocks, buy and hold is an approach that is reinforced on a monthly basis by nearly every mainstream personal finance magazine. And for years, I followed the herd, subscribed to the magazines, and tried to divine a company's true market value. Sometimes I was successful, sometimes not.

Then one day I purchased Trend Following: How Great Traders Make Millions in Up or Down Markets by Michael Covel, and suddenly fundamental analysis seemed so quaint and irrelevant.

Who cares how much a company is really worth? If its stock is going up, then buy it and go along for the ride. When it starts going down, sell. Or better yet, sell short and try to profit on the way down too! In other words, the trend is your friend, ride it and then get off.

The concept of Trend Following is so clear. So uncluttered. So basic. Afterall, what's more basic than making trading decisions based upon just the price? Or price plus volume? Or price plus volume plus an indicator or two?

As I soon found out, the concept is clear, basic and uncluttered, but the devil is in the details, such as, how does one define a trend? And once you have settled on a definition for a trend, how do you program that definition into trading software that will issue buy and sell signals?

Of course, once the program is completed it has to be backtested to determine if it is profitable or not. Then finally, once you have succeeded in programming a winning trading strategy, you have to stick to it long enough for the system to catch its stride and begin performing to its potential. It is at this point that human emotions take over and potentially profitable systems crash and burn, unwitting victims of the human psyche.

For as long as I can remember, the stock market has been a puzzle to be studied and solved. Learning to win as a trader, I believe, is the ultimate intellectual challenge...a challenge that will play itself out in this blog.

I do love this stuff so!