Of particular interest to those who trade with signals provided by RSI(2) is the tendency of the measure to portend a reversion to the mean. When RSI(2) crosses an upper threshold, it is often a signal to close longs or to sell short. Similarly, when it crosses a lower threshold, it is often a reliable signal to cover shorts and initiate longs.

What makes trading RSI(2) exciting is that when markets begin trending, rather than reverting to the mean, RSI(2) will often stay elevated, or depressed, for extended periods of time. When it stays elevated, it appears to me that it just floats for days at a time, not really moving up or down. Unfortunately, a floating RSI(2) can be painful when one was expecting a quick move in the opposite direction. I know the last 3 days have been painful for me, as I’ve been net short the S&P 500.

I decided to take a look at any edge this setup might provide. What I tested was going long or short the SPX after RSI(2) closed above 90 for 3 days in a row. The test then closed the trade out X days later.

 I was rather embarrassed after I completed the study to realize that today the SPX just missed having its RSI(2) close above 90 for a third day in a row. RSI(2) closed at 88.07 It is probably close enough to still make the study applicable to the next few days of trading.

Results are below and cover 10 years of data for the S&P 500.

The top sheet shows the results for going long the SPX after the setup, and the bottom sheet shows the results of shorting after the setup.

It appears that the Bears have the edge over the next 8 days, with the next 3 days showing a good likelihood of a pullback.

There are a few other metrics that should be of interest. I particularly find the difference between the drawdowns to be significant. Also, the profit factor for the shorts is roughly 4x that of the long side.

Caveats:

After 10 days, the longs start working much better and show good net profits which peak near the 25th day.

Had I looked back another 30 years or so, the long trade would have out-performed the short side, even over a shorter time frame. RSI(2) did not always work well as a mean-reversion measure.

__________________________________________________________________________________________________

I included the chart of DIG as it shows an excellent example of a floating RSI(2).

After going into hibernation for over a month, the Big Bamboo is close to giving signals, on the long side.

The system has been waiting for a discernible trend in order to begin trading again. While it is not certain that the current trend will be maintained, the Bamboo is waiting for a pullback, before getting long. If it deems the pullback is orderly and not too brutal, it will start trading again.

Personally, I can hardly wait to see how it performs in an uptrend.

In the post The Return of Strategy Trading I presented an entry and exit setup from a system that performed reasonably well throughout 2008.

The chart above shows the entry day. The rules of the system dictate closing the trade tomorrow, on the open. As this stock can open without a great amount of liquidity, I might suggest holding off a minute or two before placing the sell order. A limit order might work as well.

If one was able to get in MAXY as suggested at the open on Wednesday, December 31st, and can exit tomorrow near $9.00, the percentage gain will be roughly 5.5%

After finishing my accounting for 2008, I will have to amend my numbers a tad from where I reported them in Part 1.  Take the report for what it is: not audited. The results include all commissions, charges, and platform fees. 

In May, I split my account in half, using one-half for only system trades and the other half to trade as I see fit. In 2009, these accounts will again be combined to only trade mechanical systems.

What Worked in 2008

• I will always remember 2008 as the year of Reversion to the Mean. The system that I traded almost exclusively (except for 3 trades) in my system account is a mean reversion system. The metrics posted in Part 1 are a testament to how well these types of systems performed.
• Strategy Development: With a lot of my own work and the help of other bloggers such as Damian, Michael, Brandon, and Rob, I was able to develop 6 mechanical systems. These systems are all due to be deployed in 2009.
• Covestor: Registering with Covestor was an important development as it allowed me to demonstrate (we’ll call it auditing-lite) the performance of one of the systems as well as increase public interest in my blog and the systems.
• Rhythm: In 2008 I finally hit a nice rhythm where I was able to balance the demands of working full time, blogging, trading and developing systems, and having a family. Credit to that goes to trading mechanically for most of the year as it required less time and emotional input than a discretionary approach.
• Volatility Stops: I’m not going to expound on them right now, but these are the ultimate survival tool in markets like those we saw in the first three and last three months of 2008.
• Position-Sizing: Again, no need to expound here, but effective position-sizing was the other key to surviving (and profiting) from 2008.

What I Could Have Done Better in 2008

• More Risk!!! I should have taken on twice as much risk per trade. Almost all trades in 2008 were sized to lose .5% of total capital. In hindsight, it seems ridiculous, with a system that was right more than 75% of the time, that I didn’t risk more in each trade.
• More Opportunity: As the indexes dislocated themselves daily, the systems began providing fewer and fewer opportunities. As Opportunity x Expectancy = Profit, fewer opportunities, coupled with too little risk, resulted in performance that was 1/3rd of what it could have been. In hindsight, I should have loosened the parameters of some the systems. This would have theoretically lowered expectancy. However, expectancy would still have been positive, and more opportunities in the volatile environment would likely have made up for the decrease in performance.
• Lower Commissions: Unless you are trading a size that benefits from a flat fee per trade, per share commissions are the only way to go. I will not go through 2009 paying a flat fee per trade. Per trade commissions alone subtracted better than 2% from my total return.
• Stupid Discretionary Trades: Even as I realized that I was failing as a discretionary trader, I continued to make some discretionary trades. This was not healthy for the account and has been discontinued for the foreseeable future.

New for 2009

• Full 3rd Party Auditing on all master accounts and likely on individual strategies.
• Better and more comprehensivetesting of ideas on the blog. I’m looking forward to having more time to spend looking at the strategies that other bloggers such as RipeTrade and BZB write about.
• More live system tracking, similar to what I’m doing with The Big Bamboo.
• Developments in scalability on some of the strategies.
• A formal partnership and trading business, Algorithmic Capital, LLC. (More on this in the near future).

I have a distinct feeling that 2009 is going to be a very good year. To all the iBC bloggers and everyone else, I wish you the best and most prosperous year yet.

Really, this is all about making money trading. Therefore, Part 1 will detail my trading profits.

In Part 2 I will discuss what worked and what could have worked better. I will also discuss the changes I will make in 2009.

Here is a quick summary. My strategy account was up 11.53% (19.16% annualized), while my discretionary account finished flat. Combining the two accounts brings my overall gains in 2008 to 6.03%.  I’m not happy with my lackluster performance in my discretionary account, but that is more a conversation for Part 2.

I want to focus on the metrics of the strategy trades. The performance statistics are below. I find them nothing less than stellar. The metrics that I found especially appealing are highlighted in green.

Note the montly performance report. I had only one down month equaling a percentage loss of -0.39%. To be clear, the account made its first trade on May 19th, so I was down 1 out of 8 possible months, during the worst market since the Great Depression.

Feel free to leave any questions about my performance or the report in the comments section.

I want to wish a Happy and Prosperous New Year to the iBC community. I also want to issue a sincere thank you to all of the readers who have supported me over the years. 2008 has been a pivotal year for me, and it absolutely could not have happened without the blog and the readers.

After months of dislocation, moving averages are beginning to catch up with equities. This is a good thing as most of the strategies I trade have at least one moving average as an integral component.

One of the systems that has performed reasonably well this year (46.82% annualized ROI; 56.6% win rate; .53% average net change/trade) is the Power Dip. This is a classic dip buying strategy. It is only because of sound money/risk management that this long-only system has survived the great bear of 2008.

The last signal issued by the Power Dip was in late September. Today the system was very very close to giving a signal for MAXY, for a buy on tomorrow’s open. Technically, the strong close today means MAXY didn’t make the cut, but the difference between triggering the entry and not is neglible. Therefore I still like the setup.

Following the rules will require using an 8% stop with an exit the morning after an RSI2 close over 80. The stock can be held for a maximum of 3 days, and then it must be sold.

Caveat: I believe MAXY is a biotech. Beware that biotechs have an affinity for blowing off arms and legs of those who trade them.

Starting equity was $4,000 which was enough to buy 100 shares of the Qs on 1-5-04, the date the first trade was made.

The rules are simple. Buy 100 shares at the close if the close is below the MA(5). Sell 100 shares at the close if the close is above the MA(5). The system simultaneously opens a short position on any close above the MA(5) and closes the short position on any close beneath the MA(5). Thus, the system is invested 100% of the time, either long or short.

Money management rules: The system will add or subtract the equity from any gain or loss to the 100 share size. After a winning trade, the next trade may purchase 101 shares, or after a loss, 99 shares.  This means that the system is reinvesting profits or is trading smaller after losses. A refinement to the system would have the longs and shorts tracked separately so that as the shorts were working, their size gets larger while long sizes are decreased.

There are no stops used. I tested a time-based exit and suprisingly it worsened some drawdowns. I have not tested a percentage or volatility-based stop.

Commissions were calculated at .01/share. The report shows that commissions totaled over 25% of the starting equity.

I did not factor in slippage as I think this system may be an excellent candidate to trade with EOD funds.

Caveats: This system has performed very poorly, for many many years. While a different moving average would had eeked out some profits in the 1990s, the system did not begin performing very well until 2003. Keep in mind that this report ignores the years when it did not work very well.

My suspicion is that this system does a very good job of capturing volatility. Thus, it may have performed well during other periods of high volatility, such as the early 1970s, but I have not tested it on that period exclusively.

The Equity Curve

______________________________________________________________________________________________

The screen shot below includes the MA(5) and the longs and shorts from September to the present. The largest intraday drawdown of the system was incurred on October 10th, 2008. Note the system goes long 291 shares on September 29th. It exits on October 13th after the Qs gapped above the MA(5).

Also note that one of the most profitable shorts of the 5 year period was initiated on October 13th, after the largest trade to trade drawdown was booked.

I want to thank El Cuervo for bringing this system to our attention. I think it has the potential to develop into something that I may add to my personal system quiver. I’ll consider any suggestions for improvement in the comments section, and will test them (given enough time and assuming I can code the changes) and report back any success.

For the rules of this system, check out Cuervo’s post where he explains his Qsinator 5 Day Moving Average Trading System.

Starting equity was $5,000 with 100 shares traded each time.

I included .01 share for commission.

You’ll find my results (below) to be similar to his. My testing used the closing price of the day that price crossed above or beneath the 5 day simple moving average for entry and exit. Cuervo required the high of the day to be beneath the moving average to trigger an entry. I did not code it with that additional requirement.

I’ve circled in green the metrics that I pay the most attention to, but there are others worth reviewing such as the Profit Factor and Ratio Avg. Win:Avg. Loss. Also note that the average losing trade lasted twice as long as the average winning trade. You might be quizzed on that later.

________________________________________________________________________________________________

Below is the equity curve. Note the two large drawdowns. The first large drawdown began on 12-27-07 as the system went long twice during the January swoon. The low of this drawdown was January 23, where the system had lost 16% (intraday) from starting equity. From trade start to finish, the drawdown ended up being 9.5%

The second large drawdown began at the end of September. By October 10th, the system had lost (intraday)over 16% from the high at the end of September. However, from trade start to finish, the drawdown was only 6.5%

I believe there is a very simple refinement that could be made to decrease the drawdown. Anyone have a guess what that refinement could be?

________________________________________________________________________________________________

For fun, and because it looks cool, here are some of the recent trades the system has taken. Per El Cuervo’s rules, the system only traded 100 shares at a time. On the chart below, MA5LE = Long Entry while MA5LX = Long Exit.

A very simple yet profitable system. I wonder what percentage of retail traders were able to beat this system in 2008?

Caveat: Testing over all the data for the Qs (starts March 1999), the system hit a high in March of 2000 that was not surpassed until August 2008.