Power Dip Testing: Uptrends, Volatility, Bear Markets

A summary is below the sheet.

Click on spreadsheet to enlarge…

November will mark the one year anniversary of the inception of PDS subscription service. As we approach the beginning of the second year, I am in the process of reviewing the system in order to make any necessary changes for implementation the 2nd year. I’m not anticipating any changes to the original algorithm; rather, the changes will likely involve an adaptive position-sizing method, or at the least, refining the original methods.

The tests represented above attempt to answer two questions:

1. Would adding an additional uptrend requirement and trading only low or high volatility stocks improve the system?

2. Does ATR position-sizing or fixed-stop position-sizing work better during bear markets?

Reading the Sheet:

All tests include .01/share for commissions and were run over de-listed data from Premium Data.

• The top half of the sheet (above the blue horizontal line) shows the results of testing from 7/1/2007 – 9/4/2010. I consider this period to encompass the current bear market (which includes the bull run of 2009).
• The bottom half of the sheet (below the blue horizontal line) shows the results of testing from 4/1/2000 – 6/1/2003. This period represents the bear market that followed the tech crash.
• The white columns show the fixed-stop results (1% Risk, 10% Stop) with the uptrend requirement (U) and either low (L) or high (H) volatility stocks.
• The gray columns show the same tests with ATR position-sizing using 1% Risk.

Interpreting the Results:

For this type of system, opportunity is everything. I cannot emphasize that enough.

• While the U_H (uptrend and high volatility) models have a significantly higher average trade during the 2000 bear market period than the baseline models, opportunity is limited to the extent that they under perform in terms of net percentage gain. However, the higher average trade is encouraging and it may mean that we flag those higher volatility picks so that subscribers can adjust position-sizing in an attempt to capture a larger gain. The data show other added benefits to the U_H and U_L picks.
• During bear markets, ATR position-sizing appears to under perform fixed-stop position-sizing. This is contrary to my own intuition. While I need to run similar tests during bull markets to be sure, I’m fairly sure that ATR position-sizing is most beneficial when volatility is low (as we would expect during bull markets) as it requires taking larger positions than we would with fixed-stop sizing. Excluding net percentage gain, the data show that there are other benefits to using ATR position-sizing rather than fixed-stop.

More Testing Ahead…

I am currently working on an adaptive model which will change both risked amount and position-sizing models based on the market environment. Also, I have combined ATR position-sizing with fixed-stops, and the results are encouraging. I will publish the results as soon as they are available.

Comments »

Plug these in to your favorite charting program and take a look.

DECK

EWBC

BLC

PEI

CAAS

JDSU

OCLR

BC

FST

CRUS

Comments »

August was a horrible month for equities while PDS mostly dodged the downturn.

A note about how the report is generated. The monthly statistics, %P/L and %CAR are calculated from January 4th (first trading day of 2010) to the present, NOT starting from August 2nd to August 31st. The difference is that the system may have been holding open positions going into August that affect that months performance where if one started trading the system on August 2nd (first trading day of August), there wouldn’t have been any already opened positions.

The rest of the statistics, W/L%, Avg.Trade, and Trades are calculated from August 2nd to August 31st in order to give an accurate account of one month’s performance.

Year-To-Date Performance as of 9/1/10:

YTD PDS is outperforming, but not by a wide margin. As the S&P500 has spent about half of 2010 beneath the 50 day moving average, this market is challenging for a long-only system. Furthermore, PDS requires stocks be in an uptrend in order to be considered. There simply haven’t been many stocks in an uptrend due to the market weakness, which means less opportunity for the system to profit (note the exposure percentages which show that the system has spent a lot of the year with cash not invested).

For those not familiar with the system, the different models (1% Risk, 10% Stop) do not represent different rules for the system but rather different methods for betting on its stock picks.

The system closed its last open position today (for a win) and is awaiting a pullback in the broader markets to generate some dips.

Comments »

I have too many irons in the fire.

I am going to take a short hiatus from posting here (just a few days or so) while I work on some money management / position-sizing tests for PDS. I am also testing 2 additional factors on the original algorithm. Subscribers can expect some updates on the Power Dip site blog.

The other major project I am working on is the proper code for the Momentum System in order to account for un-adjusted prices and volume when working with de-listed data.

Both endeavors are extremely important. Fortunately, the work on de-listed data with the Momentum System will be posted here, to the benefit of everyone who is interested. Unfortunately, unless you are a subscriber to PDS, you will not be privy to most of the work being done there.

Back soon!

Comments »

Here is the equity curve from a simple mean-reversion system. How simple is it? Very, very simple.  It is RSI2 based with a couple other factors thrown in that regulate position-sizing to limit risk during times when the market is behaving abnormally and applies 2x leverage when trading with the trend.  Anyway, after the 2008 disaster (which was a boon for the system), it has not done very much. However, it has recently begun to perform very well.

This could have several implications about the future, and we know that mean-reversion performs better during elevated volatility.

MarketSci’s last post on mean-reversion also seemed to suggest that we might see an uptick.

As you were…

Comments »

On the subscriber site, there is a fair amount of discussion about stops and position-sizing. The two issues actually go hand-in-hand. I want to address this relationship between stops and position-sizing.

This assumes that a trader uses percent-risk position-sizing. Percent-risk position-sizing means that one risks X% of his account on each trade. In order to manage the amount of money risked, we must know when a trade will be exited so that the account loses the specified amount. A stop-loss is the typical exit method used to ensure that trade risks only X% of the account.

Let’s look at an example.

• $10,000 account, and the trader wants to risk 1% of the account on each trade ($100.00 risked per trade).
• XYZ stock trades for $10.00/share.
• The trader could use a stop 10% beneath his entry: $10.00*.10=$100.00
• In the example above, he would have purchased 100 shares. 100 shares @ $10.00/share = $1000.00 position.
• When this $1000.00 position stops out at 10% loss, he loses his risk of $100.00

However, what if the trader wants to use a stop of 5%? Then his position would be twice as large as using a 10% stop.

• He would have purchased 200 shares. 200 shares @ $10.00/share = $2000.00 position.
• When this $2000.00 position stops out at 5% loss, he loses his risk of $100.00 ($2000.00*.05=$100.00)

So there is a direct connection between the stop-loss level and the size of the position.

With a winning setup and exit strategy, percent-risk position-sizing has multiple, extremely important implications. Hopefully some of these implications will become apparent. If not, I’ll have to do a better job later fleshing it all out.

Let’s look at a real life example.

We’ll use the Power Dip, which uses entry and exit strategies that generate a positive expectancy. This first test used a 5% stop. The tests that will follow will use larger stops. In the end, we’ll draw some conclusions about the effect the stop (and by default, the size of the position) had on system performance

Here are the stats of the backtest:

The most important stats in terms of this discussion are as follows:

• Avg. Profit/Loss %
• Winners
• Avg. Profit %
• Avg. Loss %
• Risk-Reward Ratio
• Number of Trades

As the stop-loss percentage gets smaller, we’ll see a decrease in the Avg. Loss % and an increase in the Risk-Reward Ratio. We will also see a decreased Avg. Profit/Loss % and a decreased Winners %. So with a decrease in the Avg. Profit/Loss % and decreased Winners %, how does the system make money? Good question…Since the stop is tight at 5%, we have fewer winning trades (59.22% in this example), but the losing trades lose smaller amounts (-4.82%). Despite the fact that the Avg. Profit % (4.84%) and Avg. Loss % are almost equal, the system profits because it has more winning trades than losing trades. The 5% stop keeps the average % losses about the same size as the average % gains.

Profit Distribution of Trades:

In the above graph, the large red spike shows that approximately 923 trades stopped-out. Since there were 2857 trades, this means that approximately 32% of trades stopped-out.

As we increase our stop-loss percentage, we can expect the Winner % to increase and the Avg. Loss % to increase. We will also expect the number of trades to grow. Thus, even though the system will lose more per trade on average, there will be an increased number of trades that win, AND, there will be increased opportunity to trade (remember, as our stop loss widens, our position-size will decrease, meaning we will have more cash available in the portfolio for more positions). Increased opportunity means more chance for profits.

Now, the Obligatory Equity Curve:

Notice that a 5% stop works well when volatility is low (duh?).

Over the next couple of days, I’ll adjust the stop level upwards, and run the same test. By comparing the metrics, we’ll begin to get a sense of the effect of stop-loss and position-sizing on a positive expectancy system.

Your comments are welcome.

Comments »

When adding de-listed data to the momentum system, performance degrades significantly, even over a short time horizon of one-month. I would expect to see performance degrade with longer minimum hold times such as six months or a year, but I am really surprised how much de-listed data affects even short-term holds.

I am re-running the last test of the momentum system with de-listed data and new code to account for stock splits, and while the test will not be done until late tonight, I can already see that performance will be degraded. I will post the results when they are ready so that we can observe how using unadjusted data affected the results vs. the adjusted data. (To see the impetus for the change of code and the thinking behind it, go here. Be sure to read the comments after the post.)

But the question remains: Why is de-listed data have such a profound affect on the system, even over short holds of 22 bars?

I am really curious to hear everyone’s thoughts on this issue.

Comments »